{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9994236311239193,
  "eval_steps": 500,
  "global_step": 867,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.050720461095100866,
      "grad_norm": 4.101463794708252,
      "learning_rate": 5.057471264367817e-07,
      "logits/chosen": -0.5009022951126099,
      "logits/rejected": -0.4338029623031616,
      "logps/chosen": -68.60438537597656,
      "logps/rejected": -11.815346717834473,
      "loss": 0.6943,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": -0.0008158835698850453,
      "rewards/margins": -0.0018830103799700737,
      "rewards/rejected": 0.0010671267518773675,
      "step": 44
    },
    {
      "epoch": 0.10144092219020173,
      "grad_norm": 5.229926109313965,
      "learning_rate": 9.987179487179487e-07,
      "logits/chosen": -0.5297297239303589,
      "logits/rejected": -0.46696794033050537,
      "logps/chosen": -62.21626663208008,
      "logps/rejected": -10.197611808776855,
      "loss": 0.6841,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": 0.013842731714248657,
      "rewards/margins": 0.01867145486176014,
      "rewards/rejected": -0.004828720819205046,
      "step": 88
    },
    {
      "epoch": 0.1521613832853026,
      "grad_norm": 6.2486772537231445,
      "learning_rate": 9.423076923076923e-07,
      "logits/chosen": -0.5430326461791992,
      "logits/rejected": -0.4660782814025879,
      "logps/chosen": -66.80805206298828,
      "logps/rejected": -11.131633758544922,
      "loss": 0.6512,
      "rewards/accuracies": 0.917613685131073,
      "rewards/chosen": 0.07174928486347198,
      "rewards/margins": 0.08846410363912582,
      "rewards/rejected": -0.01671481691300869,
      "step": 132
    },
    {
      "epoch": 0.20288184438040346,
      "grad_norm": 4.249594688415527,
      "learning_rate": 8.858974358974358e-07,
      "logits/chosen": -0.5167344212532043,
      "logits/rejected": -0.4457302987575531,
      "logps/chosen": -64.10420227050781,
      "logps/rejected": -12.427456855773926,
      "loss": 0.5764,
      "rewards/accuracies": 0.9630681872367859,
      "rewards/chosen": 0.20728658139705658,
      "rewards/margins": 0.26457884907722473,
      "rewards/rejected": -0.05729226395487785,
      "step": 176
    },
    {
      "epoch": 0.25360230547550433,
      "grad_norm": 4.136184215545654,
      "learning_rate": 8.294871794871795e-07,
      "logits/chosen": -0.5318545699119568,
      "logits/rejected": -0.45112499594688416,
      "logps/chosen": -73.85310363769531,
      "logps/rejected": -14.750091552734375,
      "loss": 0.4526,
      "rewards/accuracies": 0.9715909361839294,
      "rewards/chosen": 0.4944852590560913,
      "rewards/margins": 0.6303640007972717,
      "rewards/rejected": -0.1358787715435028,
      "step": 220
    },
    {
      "epoch": 0.3043227665706052,
      "grad_norm": 2.5871806144714355,
      "learning_rate": 7.730769230769231e-07,
      "logits/chosen": -0.5192740559577942,
      "logits/rejected": -0.44760316610336304,
      "logps/chosen": -67.3663558959961,
      "logps/rejected": -14.041060447692871,
      "loss": 0.3581,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 0.7611348032951355,
      "rewards/margins": 1.0298521518707275,
      "rewards/rejected": -0.26871728897094727,
      "step": 264
    },
    {
      "epoch": 0.35504322766570606,
      "grad_norm": 2.5204851627349854,
      "learning_rate": 7.166666666666667e-07,
      "logits/chosen": -0.5485614538192749,
      "logits/rejected": -0.48719319701194763,
      "logps/chosen": -63.37010192871094,
      "logps/rejected": -17.727149963378906,
      "loss": 0.2842,
      "rewards/accuracies": 0.9744318723678589,
      "rewards/chosen": 0.9588545560836792,
      "rewards/margins": 1.4247877597808838,
      "rewards/rejected": -0.46593329310417175,
      "step": 308
    },
    {
      "epoch": 0.40576368876080693,
      "grad_norm": 3.2390079498291016,
      "learning_rate": 6.602564102564102e-07,
      "logits/chosen": -0.5195890069007874,
      "logits/rejected": -0.471608430147171,
      "logps/chosen": -57.76165008544922,
      "logps/rejected": -19.150711059570312,
      "loss": 0.2386,
      "rewards/accuracies": 0.9659091234207153,
      "rewards/chosen": 1.1420756578445435,
      "rewards/margins": 1.7942817211151123,
      "rewards/rejected": -0.6522059440612793,
      "step": 352
    },
    {
      "epoch": 0.4564841498559078,
      "grad_norm": 1.46431303024292,
      "learning_rate": 6.038461538461538e-07,
      "logits/chosen": -0.5360719561576843,
      "logits/rejected": -0.4885927736759186,
      "logps/chosen": -59.30861282348633,
      "logps/rejected": -21.8256778717041,
      "loss": 0.1872,
      "rewards/accuracies": 0.980113685131073,
      "rewards/chosen": 1.269405484199524,
      "rewards/margins": 2.136420965194702,
      "rewards/rejected": -0.8670153617858887,
      "step": 396
    },
    {
      "epoch": 0.5072046109510087,
      "grad_norm": 1.0862195491790771,
      "learning_rate": 5.474358974358975e-07,
      "logits/chosen": -0.5177943706512451,
      "logits/rejected": -0.4745483100414276,
      "logps/chosen": -51.07387161254883,
      "logps/rejected": -20.606046676635742,
      "loss": 0.1813,
      "rewards/accuracies": 0.9772727489471436,
      "rewards/chosen": 1.2671505212783813,
      "rewards/margins": 2.2759194374084473,
      "rewards/rejected": -1.008769154548645,
      "step": 440
    },
    {
      "epoch": 0.5579250720461095,
      "grad_norm": 2.516627311706543,
      "learning_rate": 4.910256410256409e-07,
      "logits/chosen": -0.4945791959762573,
      "logits/rejected": -0.4570486843585968,
      "logps/chosen": -50.394466400146484,
      "logps/rejected": -22.478759765625,
      "loss": 0.1446,
      "rewards/accuracies": 0.9829545617103577,
      "rewards/chosen": 1.2443382740020752,
      "rewards/margins": 2.46410870552063,
      "rewards/rejected": -1.2197703123092651,
      "step": 484
    },
    {
      "epoch": 0.6086455331412104,
      "grad_norm": 1.2264533042907715,
      "learning_rate": 4.346153846153846e-07,
      "logits/chosen": -0.49034449458122253,
      "logits/rejected": -0.44115787744522095,
      "logps/chosen": -59.09516143798828,
      "logps/rejected": -25.836565017700195,
      "loss": 0.1132,
      "rewards/accuracies": 0.9886363744735718,
      "rewards/chosen": 1.4085954427719116,
      "rewards/margins": 2.830648422241211,
      "rewards/rejected": -1.4220527410507202,
      "step": 528
    },
    {
      "epoch": 0.6593659942363113,
      "grad_norm": 1.2289067506790161,
      "learning_rate": 3.782051282051282e-07,
      "logits/chosen": -0.5277236104011536,
      "logits/rejected": -0.48551374673843384,
      "logps/chosen": -54.49405288696289,
      "logps/rejected": -26.42508888244629,
      "loss": 0.1041,
      "rewards/accuracies": 0.9886363744735718,
      "rewards/chosen": 1.3465783596038818,
      "rewards/margins": 2.97350811958313,
      "rewards/rejected": -1.6269298791885376,
      "step": 572
    },
    {
      "epoch": 0.7100864553314121,
      "grad_norm": 2.022843599319458,
      "learning_rate": 3.217948717948718e-07,
      "logits/chosen": -0.5311535000801086,
      "logits/rejected": -0.47788894176483154,
      "logps/chosen": -54.31318283081055,
      "logps/rejected": -26.987857818603516,
      "loss": 0.095,
      "rewards/accuracies": 0.9857954978942871,
      "rewards/chosen": 1.3066821098327637,
      "rewards/margins": 3.0658864974975586,
      "rewards/rejected": -1.759204387664795,
      "step": 616
    },
    {
      "epoch": 0.760806916426513,
      "grad_norm": 1.6733691692352295,
      "learning_rate": 2.653846153846154e-07,
      "logits/chosen": -0.5272084474563599,
      "logits/rejected": -0.48467957973480225,
      "logps/chosen": -53.461082458496094,
      "logps/rejected": -29.292964935302734,
      "loss": 0.0878,
      "rewards/accuracies": 0.9857954978942871,
      "rewards/chosen": 1.4012432098388672,
      "rewards/margins": 3.2184605598449707,
      "rewards/rejected": -1.817217469215393,
      "step": 660
    },
    {
      "epoch": 0.8115273775216139,
      "grad_norm": 1.2542502880096436,
      "learning_rate": 2.0897435897435897e-07,
      "logits/chosen": -0.5218697190284729,
      "logits/rejected": -0.48044687509536743,
      "logps/chosen": -51.774375915527344,
      "logps/rejected": -32.40013122558594,
      "loss": 0.0715,
      "rewards/accuracies": 0.9857954978942871,
      "rewards/chosen": 1.412352204322815,
      "rewards/margins": 3.4855575561523438,
      "rewards/rejected": -2.0732057094573975,
      "step": 704
    },
    {
      "epoch": 0.8622478386167147,
      "grad_norm": 1.5797039270401,
      "learning_rate": 1.5256410256410257e-07,
      "logits/chosen": -0.5329575538635254,
      "logits/rejected": -0.489034503698349,
      "logps/chosen": -51.42525100708008,
      "logps/rejected": -31.246944427490234,
      "loss": 0.0837,
      "rewards/accuracies": 0.9715909361839294,
      "rewards/chosen": 1.3154159784317017,
      "rewards/margins": 3.3884437084198,
      "rewards/rejected": -2.0730276107788086,
      "step": 748
    },
    {
      "epoch": 0.9129682997118156,
      "grad_norm": 1.8189433813095093,
      "learning_rate": 9.615384615384616e-08,
      "logits/chosen": -0.5260158777236938,
      "logits/rejected": -0.4700179696083069,
      "logps/chosen": -59.344993591308594,
      "logps/rejected": -32.7772102355957,
      "loss": 0.0751,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 1.5358730554580688,
      "rewards/margins": 3.563418388366699,
      "rewards/rejected": -2.027545213699341,
      "step": 792
    },
    {
      "epoch": 0.9636887608069165,
      "grad_norm": 1.5270038843154907,
      "learning_rate": 3.9743589743589737e-08,
      "logits/chosen": -0.5239982008934021,
      "logits/rejected": -0.46047428250312805,
      "logps/chosen": -61.70787048339844,
      "logps/rejected": -34.90862274169922,
      "loss": 0.0653,
      "rewards/accuracies": 0.980113685131073,
      "rewards/chosen": 1.5698310136795044,
      "rewards/margins": 3.659027099609375,
      "rewards/rejected": -2.0891964435577393,
      "step": 836
    },
    {
      "epoch": 0.9994236311239193,
      "step": 867,
      "total_flos": 5.2580366127857664e+17,
      "train_loss": 0.2640415610189119,
      "train_runtime": 13543.6511,
      "train_samples_per_second": 0.512,
      "train_steps_per_second": 0.064
    }
  ],
  "logging_steps": 44,
  "max_steps": 867,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.2580366127857664e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}