{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.138086062941555,
  "eval_steps": 500,
  "global_step": 1000,
  "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
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.10757168382406235,
      "learning_rate": 0.00040391536883141455,
      "logits/chosen": -3.046875,
      "logits/rejected": -3.046875,
      "logps/chosen": -306.0,
      "logps/rejected": -544.0,
      "loss": 0.0201,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.99609375,
      "rewards/chosen": -3.078125,
      "rewards/margins": 17.875,
      "rewards/rejected": -21.0,
      "step": 360,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.026544878259301186,
      "learning_rate": 0.0003969463130731183,
      "logits/chosen": -2.953125,
      "logits/rejected": -2.921875,
      "logps/chosen": -304.0,
      "logps/rejected": -548.0,
      "loss": 0.0068,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.99609375,
      "rewards/chosen": -3.171875,
      "rewards/margins": 18.625,
      "rewards/rejected": -21.75,
      "step": 370,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.012133469805121422,
      "learning_rate": 0.0003897982258676867,
      "logits/chosen": -2.875,
      "logits/rejected": -2.828125,
      "logps/chosen": -316.0,
      "logps/rejected": -540.0,
      "loss": 0.0172,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.99609375,
      "rewards/chosen": -3.203125,
      "rewards/margins": 18.0,
      "rewards/rejected": -21.25,
      "step": 380,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.05422056466341019,
      "learning_rate": 0.00038247981605830125,
      "logits/chosen": -2.9375,
      "logits/rejected": -2.859375,
      "logps/chosen": -306.0,
      "logps/rejected": -548.0,
      "loss": 0.0067,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.99609375,
      "rewards/chosen": -3.25,
      "rewards/margins": 19.875,
      "rewards/rejected": -23.125,
      "step": 390,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.5953335762023926,
      "learning_rate": 0.000375,
      "logits/chosen": -3.0625,
      "logits/rejected": -3.015625,
      "logps/chosen": -300.0,
      "logps/rejected": -588.0,
      "loss": 0.0196,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.99609375,
      "rewards/chosen": -3.6875,
      "rewards/margins": 21.625,
      "rewards/rejected": -25.25,
      "step": 400,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.11,
      "grad_norm": 4.781039714813232,
      "learning_rate": 0.0003673678906964727,
      "logits/chosen": -3.015625,
      "logits/rejected": -3.0,
      "logps/chosen": -300.0,
      "logps/rejected": -564.0,
      "loss": 0.0067,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.99609375,
      "rewards/chosen": -3.03125,
      "rewards/margins": 19.25,
      "rewards/rejected": -22.375,
      "step": 410,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.0018408839823678136,
      "learning_rate": 0.00035959278669726934,
      "logits/chosen": -2.953125,
      "logits/rejected": -2.921875,
      "logps/chosen": -300.0,
      "logps/rejected": -600.0,
      "loss": 0.0012,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.5625,
      "rewards/margins": 22.75,
      "rewards/rejected": -26.25,
      "step": 420,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.0082088652998209,
      "learning_rate": 0.0003516841607689501,
      "logits/chosen": -3.0625,
      "logits/rejected": -3.015625,
      "logps/chosen": -308.0,
      "logps/rejected": -560.0,
      "loss": 0.0058,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.99609375,
      "rewards/chosen": -2.828125,
      "rewards/margins": 21.0,
      "rewards/rejected": -23.875,
      "step": 430,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.0047104801051318645,
      "learning_rate": 0.00034365164835397803,
      "logits/chosen": -3.15625,
      "logits/rejected": -3.140625,
      "logps/chosen": -290.0,
      "logps/rejected": -536.0,
      "loss": 0.0025,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -2.171875,
      "rewards/margins": 18.375,
      "rewards/rejected": -20.5,
      "step": 440,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.10709760338068008,
      "learning_rate": 0.0003355050358314172,
      "logits/chosen": -3.171875,
      "logits/rejected": -3.140625,
      "logps/chosen": -310.0,
      "logps/rejected": -560.0,
      "loss": 0.0015,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -2.75,
      "rewards/margins": 21.25,
      "rewards/rejected": -24.0,
      "step": 450,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.016811428591609,
      "learning_rate": 0.00032725424859373687,
      "logits/chosen": -3.125,
      "logits/rejected": -3.078125,
      "logps/chosen": -304.0,
      "logps/rejected": -596.0,
      "loss": 0.0003,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.40625,
      "rewards/margins": 22.875,
      "rewards/rejected": -26.25,
      "step": 460,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.05604538321495056,
      "learning_rate": 0.0003189093389542498,
      "logits/chosen": -3.078125,
      "logits/rejected": -3.03125,
      "logps/chosen": -318.0,
      "logps/rejected": -612.0,
      "loss": 0.0003,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.828125,
      "rewards/margins": 24.25,
      "rewards/rejected": -28.0,
      "step": 470,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.10262610763311386,
      "learning_rate": 0.0003104804738999169,
      "logits/chosen": -3.0,
      "logits/rejected": -2.9375,
      "logps/chosen": -304.0,
      "logps/rejected": -608.0,
      "loss": 0.0063,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.99609375,
      "rewards/chosen": -4.0,
      "rewards/margins": 23.125,
      "rewards/rejected": -27.125,
      "step": 480,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.52,
      "grad_norm": 2.782456874847412,
      "learning_rate": 0.0003019779227044398,
      "logits/chosen": -2.859375,
      "logits/rejected": -2.75,
      "logps/chosen": -322.0,
      "logps/rejected": -544.0,
      "loss": 0.0095,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.99609375,
      "rewards/chosen": -3.203125,
      "rewards/margins": 18.875,
      "rewards/rejected": -22.125,
      "step": 490,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.1368335485458374,
      "learning_rate": 0.00029341204441673266,
      "logits/chosen": -3.03125,
      "logits/rejected": -2.9375,
      "logps/chosen": -296.0,
      "logps/rejected": -556.0,
      "loss": 0.0029,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -2.359375,
      "rewards/margins": 19.875,
      "rewards/rejected": -22.125,
      "step": 500,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.009250438772141933,
      "learning_rate": 0.00028479327524001636,
      "logits/chosen": -3.171875,
      "logits/rejected": -3.109375,
      "logps/chosen": -318.0,
      "logps/rejected": -552.0,
      "loss": 0.0044,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -2.796875,
      "rewards/margins": 19.75,
      "rewards/rejected": -22.625,
      "step": 510,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.1493176519870758,
      "learning_rate": 0.0002761321158169134,
      "logits/chosen": -3.21875,
      "logits/rejected": -3.1875,
      "logps/chosen": -298.0,
      "logps/rejected": -524.0,
      "loss": 0.0048,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -2.46875,
      "rewards/margins": 17.5,
      "rewards/rejected": -20.0,
      "step": 520,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.06779959052801132,
      "learning_rate": 0.0002674391184360313,
      "logits/chosen": -3.234375,
      "logits/rejected": -3.21875,
      "logps/chosen": -298.0,
      "logps/rejected": -528.0,
      "loss": 0.0038,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -2.4375,
      "rewards/margins": 17.625,
      "rewards/rejected": -20.125,
      "step": 530,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.3520822525024414,
      "learning_rate": 0.0002587248741756253,
      "logits/chosen": -3.25,
      "logits/rejected": -3.1875,
      "logps/chosen": -302.0,
      "logps/rejected": -536.0,
      "loss": 0.0021,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -2.734375,
      "rewards/margins": 19.5,
      "rewards/rejected": -22.25,
      "step": 540,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.020363105461001396,
      "learning_rate": 0.00025,
      "logits/chosen": -3.15625,
      "logits/rejected": -3.125,
      "logps/chosen": -310.0,
      "logps/rejected": -572.0,
      "loss": 0.0016,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.40625,
      "rewards/margins": 21.625,
      "rewards/rejected": -25.0,
      "step": 550,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.01094722468405962,
      "learning_rate": 0.00024127512582437484,
      "logits/chosen": -3.109375,
      "logits/rejected": -3.0625,
      "logps/chosen": -316.0,
      "logps/rejected": -640.0,
      "loss": 0.0016,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.90625,
      "rewards/margins": 26.125,
      "rewards/rejected": -30.0,
      "step": 560,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.10121448338031769,
      "learning_rate": 0.00023256088156396867,
      "logits/chosen": -3.078125,
      "logits/rejected": -3.03125,
      "logps/chosen": -332.0,
      "logps/rejected": -628.0,
      "loss": 0.0016,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.875,
      "rewards/margins": 26.375,
      "rewards/rejected": -30.25,
      "step": 570,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.01434832252562046,
      "learning_rate": 0.00022386788418308668,
      "logits/chosen": -3.0,
      "logits/rejected": -2.90625,
      "logps/chosen": -328.0,
      "logps/rejected": -640.0,
      "loss": 0.0004,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -4.375,
      "rewards/margins": 26.75,
      "rewards/rejected": -31.125,
      "step": 580,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.0004497322079259902,
      "learning_rate": 0.0002152067247599837,
      "logits/chosen": -2.953125,
      "logits/rejected": -2.875,
      "logps/chosen": -304.0,
      "logps/rejected": -656.0,
      "loss": 0.0005,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -4.375,
      "rewards/margins": 28.25,
      "rewards/rejected": -32.75,
      "step": 590,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.010737881064414978,
      "learning_rate": 0.00020658795558326743,
      "logits/chosen": -2.984375,
      "logits/rejected": -2.859375,
      "logps/chosen": -330.0,
      "logps/rejected": -676.0,
      "loss": 0.0008,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -5.03125,
      "rewards/margins": 29.375,
      "rewards/rejected": -34.5,
      "step": 600,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.13,
      "grad_norm": 0.3118731379508972,
      "learning_rate": 0.0001980220772955602,
      "logits/chosen": -3.015625,
      "logits/rejected": -2.9375,
      "logps/chosen": -310.0,
      "logps/rejected": -648.0,
      "loss": 0.0011,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.828125,
      "rewards/margins": 27.75,
      "rewards/rejected": -31.5,
      "step": 610,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.19,
      "grad_norm": 0.04152989387512207,
      "learning_rate": 0.0001895195261000831,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.03125,
      "logps/chosen": -302.0,
      "logps/rejected": -604.0,
      "loss": 0.0002,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -2.75,
      "rewards/margins": 24.75,
      "rewards/rejected": -27.5,
      "step": 620,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.00012200848141219467,
      "learning_rate": 0.00018109066104575022,
      "logits/chosen": -3.125,
      "logits/rejected": -3.0625,
      "logps/chosen": -324.0,
      "logps/rejected": -584.0,
      "loss": 0.0013,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -2.875,
      "rewards/margins": 23.375,
      "rewards/rejected": -26.25,
      "step": 630,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.29,
      "grad_norm": 0.037591688334941864,
      "learning_rate": 0.00017274575140626317,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.046875,
      "logps/chosen": -308.0,
      "logps/rejected": -608.0,
      "loss": 0.0001,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -2.9375,
      "rewards/margins": 24.375,
      "rewards/rejected": -27.25,
      "step": 640,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.34,
      "grad_norm": 0.0018848556792363524,
      "learning_rate": 0.00016449496416858284,
      "logits/chosen": -3.109375,
      "logits/rejected": -3.03125,
      "logps/chosen": -308.0,
      "logps/rejected": -600.0,
      "loss": 0.0002,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.171875,
      "rewards/margins": 24.25,
      "rewards/rejected": -27.5,
      "step": 650,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.0032229481730610132,
      "learning_rate": 0.00015634835164602198,
      "logits/chosen": -3.078125,
      "logits/rejected": -3.046875,
      "logps/chosen": -312.0,
      "logps/rejected": -612.0,
      "loss": 0.0002,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.203125,
      "rewards/margins": 24.75,
      "rewards/rejected": -28.0,
      "step": 660,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.0005074635264463723,
      "learning_rate": 0.00014831583923105,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.015625,
      "logps/chosen": -316.0,
      "logps/rejected": -596.0,
      "loss": 0.0,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.203125,
      "rewards/margins": 24.625,
      "rewards/rejected": -27.875,
      "step": 670,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.49,
      "grad_norm": 0.007239128462970257,
      "learning_rate": 0.00014040721330273062,
      "logits/chosen": -3.109375,
      "logits/rejected": -3.046875,
      "logps/chosen": -310.0,
      "logps/rejected": -604.0,
      "loss": 0.0001,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.265625,
      "rewards/margins": 24.75,
      "rewards/rejected": -28.0,
      "step": 680,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.0034740234259516,
      "learning_rate": 0.00013263210930352737,
      "logits/chosen": -3.078125,
      "logits/rejected": -3.03125,
      "logps/chosen": -300.0,
      "logps/rejected": -624.0,
      "loss": 0.0002,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.15625,
      "rewards/margins": 25.375,
      "rewards/rejected": -28.5,
      "step": 690,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.0001397097803419456,
      "learning_rate": 0.00012500000000000006,
      "logits/chosen": -3.078125,
      "logits/rejected": -3.046875,
      "logps/chosen": -304.0,
      "logps/rejected": -620.0,
      "loss": 0.0001,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.1875,
      "rewards/margins": 24.875,
      "rewards/rejected": -28.125,
      "step": 700,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.0016049661207944155,
      "learning_rate": 0.0001175201839416988,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.03125,
      "logps/chosen": -324.0,
      "logps/rejected": -608.0,
      "loss": 0.0001,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.28125,
      "rewards/margins": 24.75,
      "rewards/rejected": -28.0,
      "step": 710,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.006374781485646963,
      "learning_rate": 0.00011020177413231333,
      "logits/chosen": -3.109375,
      "logits/rejected": -3.03125,
      "logps/chosen": -306.0,
      "logps/rejected": -608.0,
      "loss": 0.0002,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.28125,
      "rewards/margins": 25.125,
      "rewards/rejected": -28.375,
      "step": 720,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.0068184914998710155,
      "learning_rate": 0.00010305368692688174,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.046875,
      "logps/chosen": -296.0,
      "logps/rejected": -616.0,
      "loss": 0.0004,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.15625,
      "rewards/margins": 25.125,
      "rewards/rejected": -28.25,
      "step": 730,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.008304049260914326,
      "learning_rate": 9.608463116858542e-05,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.046875,
      "logps/chosen": -300.0,
      "logps/rejected": -592.0,
      "loss": 0.0002,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.046875,
      "rewards/margins": 24.0,
      "rewards/rejected": -27.0,
      "step": 740,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.03573792800307274,
      "learning_rate": 8.930309757836516e-05,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.03125,
      "logps/chosen": -296.0,
      "logps/rejected": -604.0,
      "loss": 0.0001,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.015625,
      "rewards/margins": 24.75,
      "rewards/rejected": -27.75,
      "step": 750,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.0008675418794155121,
      "learning_rate": 8.271734841028553e-05,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.03125,
      "logps/chosen": -296.0,
      "logps/rejected": -612.0,
      "loss": 0.0001,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -2.984375,
      "rewards/margins": 25.125,
      "rewards/rejected": -28.0,
      "step": 760,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 3.96,
      "grad_norm": 4.6364646550500765e-05,
      "learning_rate": 7.633540738525066e-05,
      "logits/chosen": -3.109375,
      "logits/rejected": -3.046875,
      "logps/chosen": -320.0,
      "logps/rejected": -616.0,
      "loss": 0.0001,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.25,
      "rewards/margins": 24.875,
      "rewards/rejected": -28.125,
      "step": 770,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.0028735387604683638,
      "learning_rate": 7.016504991533726e-05,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.046875,
      "logps/chosen": -304.0,
      "logps/rejected": -616.0,
      "loss": 0.0,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -2.921875,
      "rewards/margins": 25.25,
      "rewards/rejected": -28.125,
      "step": 780,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.00028936524176970124,
      "learning_rate": 6.421379363065141e-05,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.046875,
      "logps/chosen": -302.0,
      "logps/rejected": -596.0,
      "loss": 0.0001,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -2.875,
      "rewards/margins": 25.0,
      "rewards/rejected": -27.875,
      "step": 790,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.11,
      "grad_norm": 4.6098506572889164e-05,
      "learning_rate": 5.848888922025553e-05,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.046875,
      "logps/chosen": -308.0,
      "logps/rejected": -616.0,
      "loss": 0.0,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.125,
      "rewards/margins": 25.5,
      "rewards/rejected": -28.625,
      "step": 800,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.0014158826088532805,
      "learning_rate": 5.299731159831953e-05,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.046875,
      "logps/chosen": -298.0,
      "logps/rejected": -616.0,
      "loss": 0.0,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.03125,
      "rewards/margins": 25.5,
      "rewards/rejected": -28.5,
      "step": 810,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.0009728537406772375,
      "learning_rate": 4.7745751406263163e-05,
      "logits/chosen": -3.078125,
      "logits/rejected": -3.03125,
      "logps/chosen": -300.0,
      "logps/rejected": -620.0,
      "loss": 0.0001,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.125,
      "rewards/margins": 25.875,
      "rewards/rejected": -29.0,
      "step": 820,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.0001809960085665807,
      "learning_rate": 4.274060686123959e-05,
      "logits/chosen": -3.109375,
      "logits/rejected": -3.03125,
      "logps/chosen": -300.0,
      "logps/rejected": -616.0,
      "loss": 0.0,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -2.84375,
      "rewards/margins": 25.375,
      "rewards/rejected": -28.25,
      "step": 830,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.0019057235913351178,
      "learning_rate": 3.798797596089351e-05,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.046875,
      "logps/chosen": -310.0,
      "logps/rejected": -620.0,
      "loss": 0.0,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.09375,
      "rewards/margins": 25.5,
      "rewards/rejected": -28.5,
      "step": 840,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.007396820466965437,
      "learning_rate": 3.3493649053890325e-05,
      "logits/chosen": -3.078125,
      "logits/rejected": -3.015625,
      "logps/chosen": -310.0,
      "logps/rejected": -600.0,
      "loss": 0.0,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.203125,
      "rewards/margins": 24.625,
      "rewards/rejected": -27.75,
      "step": 850,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.007374617271125317,
      "learning_rate": 2.9263101785268254e-05,
      "logits/chosen": -3.078125,
      "logits/rejected": -3.03125,
      "logps/chosen": -300.0,
      "logps/rejected": -616.0,
      "loss": 0.0001,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.03125,
      "rewards/margins": 25.5,
      "rewards/rejected": -28.5,
      "step": 860,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.004346287343651056,
      "learning_rate": 2.5301488425208295e-05,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.03125,
      "logps/chosen": -310.0,
      "logps/rejected": -624.0,
      "loss": 0.0,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.359375,
      "rewards/margins": 25.5,
      "rewards/rejected": -28.75,
      "step": 870,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.0016928453696891665,
      "learning_rate": 2.1613635589349755e-05,
      "logits/chosen": -3.078125,
      "logits/rejected": -3.015625,
      "logps/chosen": -318.0,
      "logps/rejected": -620.0,
      "loss": 0.0,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.15625,
      "rewards/margins": 25.625,
      "rewards/rejected": -28.75,
      "step": 880,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.0004315236583352089,
      "learning_rate": 1.8204036358303172e-05,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.046875,
      "logps/chosen": -302.0,
      "logps/rejected": -608.0,
      "loss": 0.0,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.171875,
      "rewards/margins": 25.125,
      "rewards/rejected": -28.375,
      "step": 890,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.0036792384926229715,
      "learning_rate": 1.5076844803522921e-05,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.046875,
      "logps/chosen": -322.0,
      "logps/rejected": -624.0,
      "loss": 0.0,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.28125,
      "rewards/margins": 25.375,
      "rewards/rejected": -28.75,
      "step": 900,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.0032357031013816595,
      "learning_rate": 1.2235870926211617e-05,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.03125,
      "logps/chosen": -308.0,
      "logps/rejected": -608.0,
      "loss": 0.0001,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.0,
      "rewards/margins": 25.125,
      "rewards/rejected": -28.125,
      "step": 910,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.00263414834626019,
      "learning_rate": 9.684576015420277e-06,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.03125,
      "logps/chosen": -308.0,
      "logps/rejected": -608.0,
      "loss": 0.0001,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.265625,
      "rewards/margins": 25.375,
      "rewards/rejected": -28.625,
      "step": 920,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.0006815909291617572,
      "learning_rate": 7.426068431000882e-06,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.03125,
      "logps/chosen": -300.0,
      "logps/rejected": -612.0,
      "loss": 0.0001,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.171875,
      "rewards/margins": 25.0,
      "rewards/rejected": -28.25,
      "step": 930,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.007029928732663393,
      "learning_rate": 5.463099816548578e-06,
      "logits/chosen": -3.109375,
      "logits/rejected": -3.046875,
      "logps/chosen": -314.0,
      "logps/rejected": -624.0,
      "loss": 0.0001,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.28125,
      "rewards/margins": 26.0,
      "rewards/rejected": -29.25,
      "step": 940,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.0016696957172825933,
      "learning_rate": 3.798061746947995e-06,
      "logits/chosen": -3.078125,
      "logits/rejected": -3.03125,
      "logps/chosen": -292.0,
      "logps/rejected": -616.0,
      "loss": 0.0,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.21875,
      "rewards/margins": 25.25,
      "rewards/rejected": -28.375,
      "step": 950,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.004225951619446278,
      "learning_rate": 2.4329828146074094e-06,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.046875,
      "logps/chosen": -308.0,
      "logps/rejected": -624.0,
      "loss": 0.0,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.203125,
      "rewards/margins": 26.0,
      "rewards/rejected": -29.125,
      "step": 960,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.00014331191778182983,
      "learning_rate": 1.3695261579316775e-06,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.015625,
      "logps/chosen": -312.0,
      "logps/rejected": -604.0,
      "loss": 0.0001,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.15625,
      "rewards/margins": 25.5,
      "rewards/rejected": -28.75,
      "step": 970,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 5.04,
      "grad_norm": 0.0036776724737137556,
      "learning_rate": 6.089874350439506e-07,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.046875,
      "logps/chosen": -330.0,
      "logps/rejected": -616.0,
      "loss": 0.0001,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.3125,
      "rewards/margins": 25.25,
      "rewards/rejected": -28.5,
      "step": 980,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 5.09,
      "grad_norm": 1.2631918252736796e-05,
      "learning_rate": 1.5229324522605948e-07,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.046875,
      "logps/chosen": -300.0,
      "logps/rejected": -608.0,
      "loss": 0.0,
      "max_memory_allocated (GB)": 94.42,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.328125,
      "rewards/margins": 25.0,
      "rewards/rejected": -28.375,
      "step": 990,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 5.14,
      "grad_norm": 0.0053689004853367805,
      "learning_rate": 0.0,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.0625,
      "logps/chosen": -304.0,
      "logps/rejected": -624.0,
      "loss": 0.0,
      "max_memory_allocated (GB)": 94.42,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.265625,
      "rewards/margins": 25.25,
      "rewards/rejected": -28.625,
      "step": 1000,
      "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": 8.39920851615744e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}