{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 80.0,
  "eval_steps": 500,
  "global_step": 3040,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0526315789473684,
      "grad_norm": 8679.8642578125,
      "learning_rate": 7.312500000000001e-07,
      "loss": 1384.9168,
      "step": 40
    },
    {
      "epoch": 2.1052631578947367,
      "grad_norm": 4251.486328125,
      "learning_rate": 1.48125e-06,
      "loss": 1290.1081,
      "step": 80
    },
    {
      "epoch": 3.1578947368421053,
      "grad_norm": 4015.17138671875,
      "learning_rate": 2.23125e-06,
      "loss": 1097.6352,
      "step": 120
    },
    {
      "epoch": 4.2105263157894735,
      "grad_norm": 2110.156982421875,
      "learning_rate": 2.98125e-06,
      "loss": 846.727,
      "step": 160
    },
    {
      "epoch": 5.2631578947368425,
      "grad_norm": 1568.4326171875,
      "learning_rate": 3.73125e-06,
      "loss": 578.2698,
      "step": 200
    },
    {
      "epoch": 6.315789473684211,
      "grad_norm": 1924.275146484375,
      "learning_rate": 4.4812500000000005e-06,
      "loss": 412.4179,
      "step": 240
    },
    {
      "epoch": 7.368421052631579,
      "grad_norm": 862.5057373046875,
      "learning_rate": 5.23125e-06,
      "loss": 311.6504,
      "step": 280
    },
    {
      "epoch": 8.421052631578947,
      "grad_norm": 1463.6671142578125,
      "learning_rate": 5.98125e-06,
      "loss": 251.8019,
      "step": 320
    },
    {
      "epoch": 9.473684210526315,
      "grad_norm": 1956.26708984375,
      "learning_rate": 6.73125e-06,
      "loss": 197.8886,
      "step": 360
    },
    {
      "epoch": 10.526315789473685,
      "grad_norm": 404.4883728027344,
      "learning_rate": 7.481250000000001e-06,
      "loss": 155.3046,
      "step": 400
    },
    {
      "epoch": 11.578947368421053,
      "grad_norm": 319.11956787109375,
      "learning_rate": 8.23125e-06,
      "loss": 127.7263,
      "step": 440
    },
    {
      "epoch": 12.631578947368421,
      "grad_norm": 289.92144775390625,
      "learning_rate": 8.98125e-06,
      "loss": 101.4958,
      "step": 480
    },
    {
      "epoch": 13.68421052631579,
      "grad_norm": 238.7841033935547,
      "learning_rate": 9.731250000000001e-06,
      "loss": 80.8442,
      "step": 520
    },
    {
      "epoch": 14.736842105263158,
      "grad_norm": 194.82733154296875,
      "learning_rate": 1.048125e-05,
      "loss": 62.7129,
      "step": 560
    },
    {
      "epoch": 15.789473684210526,
      "grad_norm": 154.37205505371094,
      "learning_rate": 1.123125e-05,
      "loss": 50.3538,
      "step": 600
    },
    {
      "epoch": 16.842105263157894,
      "grad_norm": 208.374267578125,
      "learning_rate": 1.198125e-05,
      "loss": 41.4226,
      "step": 640
    },
    {
      "epoch": 17.894736842105264,
      "grad_norm": 101.1446304321289,
      "learning_rate": 1.2731250000000001e-05,
      "loss": 32.4618,
      "step": 680
    },
    {
      "epoch": 18.94736842105263,
      "grad_norm": 147.55865478515625,
      "learning_rate": 1.348125e-05,
      "loss": 27.0828,
      "step": 720
    },
    {
      "epoch": 20.0,
      "grad_norm": 112.81380462646484,
      "learning_rate": 1.423125e-05,
      "loss": 22.4964,
      "step": 760
    },
    {
      "epoch": 21.05263157894737,
      "grad_norm": 73.30062866210938,
      "learning_rate": 1.4981250000000002e-05,
      "loss": 18.7236,
      "step": 800
    },
    {
      "epoch": 22.105263157894736,
      "grad_norm": 64.0571517944336,
      "learning_rate": 1.5731250000000003e-05,
      "loss": 16.3388,
      "step": 840
    },
    {
      "epoch": 23.157894736842106,
      "grad_norm": 76.62981414794922,
      "learning_rate": 1.6481249999999997e-05,
      "loss": 14.0038,
      "step": 880
    },
    {
      "epoch": 24.210526315789473,
      "grad_norm": 83.39350891113281,
      "learning_rate": 1.723125e-05,
      "loss": 12.6697,
      "step": 920
    },
    {
      "epoch": 25.263157894736842,
      "grad_norm": 49.30060577392578,
      "learning_rate": 1.798125e-05,
      "loss": 11.5521,
      "step": 960
    },
    {
      "epoch": 26.31578947368421,
      "grad_norm": 49.78151321411133,
      "learning_rate": 1.873125e-05,
      "loss": 10.8067,
      "step": 1000
    },
    {
      "epoch": 27.36842105263158,
      "grad_norm": 64.97428894042969,
      "learning_rate": 1.9481250000000003e-05,
      "loss": 9.9502,
      "step": 1040
    },
    {
      "epoch": 28.42105263157895,
      "grad_norm": 96.64849853515625,
      "learning_rate": 2.023125e-05,
      "loss": 9.6148,
      "step": 1080
    },
    {
      "epoch": 29.473684210526315,
      "grad_norm": 65.55039978027344,
      "learning_rate": 2.098125e-05,
      "loss": 9.462,
      "step": 1120
    },
    {
      "epoch": 30.526315789473685,
      "grad_norm": 85.08692169189453,
      "learning_rate": 2.173125e-05,
      "loss": 8.8002,
      "step": 1160
    },
    {
      "epoch": 31.57894736842105,
      "grad_norm": 124.7302017211914,
      "learning_rate": 2.248125e-05,
      "loss": 8.7508,
      "step": 1200
    },
    {
      "epoch": 32.63157894736842,
      "grad_norm": 152.3477020263672,
      "learning_rate": 2.3231250000000002e-05,
      "loss": 8.5803,
      "step": 1240
    },
    {
      "epoch": 33.68421052631579,
      "grad_norm": 162.75997924804688,
      "learning_rate": 2.398125e-05,
      "loss": 8.1324,
      "step": 1280
    },
    {
      "epoch": 34.73684210526316,
      "grad_norm": 78.276611328125,
      "learning_rate": 2.4731249999999998e-05,
      "loss": 8.4976,
      "step": 1320
    },
    {
      "epoch": 35.78947368421053,
      "grad_norm": 151.76458740234375,
      "learning_rate": 2.548125e-05,
      "loss": 8.9999,
      "step": 1360
    },
    {
      "epoch": 36.8421052631579,
      "grad_norm": 67.1139144897461,
      "learning_rate": 2.623125e-05,
      "loss": 8.0035,
      "step": 1400
    },
    {
      "epoch": 37.89473684210526,
      "grad_norm": 61.43220901489258,
      "learning_rate": 2.6981250000000002e-05,
      "loss": 7.8756,
      "step": 1440
    },
    {
      "epoch": 38.94736842105263,
      "grad_norm": 68.46588134765625,
      "learning_rate": 2.773125e-05,
      "loss": 7.6236,
      "step": 1480
    },
    {
      "epoch": 40.0,
      "grad_norm": 52.86519241333008,
      "learning_rate": 2.848125e-05,
      "loss": 7.3731,
      "step": 1520
    },
    {
      "epoch": 41.05263157894737,
      "grad_norm": 32.2613525390625,
      "learning_rate": 2.923125e-05,
      "loss": 7.5533,
      "step": 1560
    },
    {
      "epoch": 42.10526315789474,
      "grad_norm": 28.07533836364746,
      "learning_rate": 2.998125e-05,
      "loss": 7.2446,
      "step": 1600
    },
    {
      "epoch": 43.1578947368421,
      "grad_norm": 51.62834167480469,
      "learning_rate": 2.91875e-05,
      "loss": 7.2844,
      "step": 1640
    },
    {
      "epoch": 44.21052631578947,
      "grad_norm": 94.51573944091797,
      "learning_rate": 2.8354166666666667e-05,
      "loss": 7.3347,
      "step": 1680
    },
    {
      "epoch": 45.26315789473684,
      "grad_norm": 33.503990173339844,
      "learning_rate": 2.7520833333333333e-05,
      "loss": 7.1209,
      "step": 1720
    },
    {
      "epoch": 46.31578947368421,
      "grad_norm": 74.56465148925781,
      "learning_rate": 2.6687499999999998e-05,
      "loss": 6.7968,
      "step": 1760
    },
    {
      "epoch": 47.36842105263158,
      "grad_norm": 32.059810638427734,
      "learning_rate": 2.5854166666666667e-05,
      "loss": 6.8781,
      "step": 1800
    },
    {
      "epoch": 48.421052631578945,
      "grad_norm": 97.38982391357422,
      "learning_rate": 2.5020833333333336e-05,
      "loss": 6.7608,
      "step": 1840
    },
    {
      "epoch": 49.473684210526315,
      "grad_norm": 34.4022102355957,
      "learning_rate": 2.4187500000000002e-05,
      "loss": 6.4979,
      "step": 1880
    },
    {
      "epoch": 50.526315789473685,
      "grad_norm": 35.79924011230469,
      "learning_rate": 2.3354166666666667e-05,
      "loss": 6.5747,
      "step": 1920
    },
    {
      "epoch": 51.578947368421055,
      "grad_norm": 100.87942504882812,
      "learning_rate": 2.2520833333333333e-05,
      "loss": 6.1757,
      "step": 1960
    },
    {
      "epoch": 52.63157894736842,
      "grad_norm": 159.83860778808594,
      "learning_rate": 2.16875e-05,
      "loss": 6.4399,
      "step": 2000
    },
    {
      "epoch": 53.68421052631579,
      "grad_norm": 58.15259552001953,
      "learning_rate": 2.0854166666666668e-05,
      "loss": 6.3292,
      "step": 2040
    },
    {
      "epoch": 54.73684210526316,
      "grad_norm": 23.384042739868164,
      "learning_rate": 2.0020833333333333e-05,
      "loss": 6.2464,
      "step": 2080
    },
    {
      "epoch": 55.78947368421053,
      "grad_norm": 22.12693977355957,
      "learning_rate": 1.91875e-05,
      "loss": 6.1714,
      "step": 2120
    },
    {
      "epoch": 56.8421052631579,
      "grad_norm": 17.758420944213867,
      "learning_rate": 1.8354166666666668e-05,
      "loss": 5.9527,
      "step": 2160
    },
    {
      "epoch": 57.89473684210526,
      "grad_norm": 39.71998977661133,
      "learning_rate": 1.7520833333333333e-05,
      "loss": 5.9018,
      "step": 2200
    },
    {
      "epoch": 58.94736842105263,
      "grad_norm": 28.227684020996094,
      "learning_rate": 1.6687500000000002e-05,
      "loss": 5.8687,
      "step": 2240
    },
    {
      "epoch": 60.0,
      "grad_norm": 18.77845573425293,
      "learning_rate": 1.5854166666666668e-05,
      "loss": 5.7978,
      "step": 2280
    },
    {
      "epoch": 61.05263157894737,
      "grad_norm": 25.35810089111328,
      "learning_rate": 1.5020833333333334e-05,
      "loss": 5.8052,
      "step": 2320
    },
    {
      "epoch": 62.10526315789474,
      "grad_norm": 92.96112823486328,
      "learning_rate": 1.41875e-05,
      "loss": 5.7386,
      "step": 2360
    },
    {
      "epoch": 63.1578947368421,
      "grad_norm": 35.39582824707031,
      "learning_rate": 1.3354166666666667e-05,
      "loss": 5.6511,
      "step": 2400
    },
    {
      "epoch": 64.21052631578948,
      "grad_norm": 16.94559669494629,
      "learning_rate": 1.2520833333333334e-05,
      "loss": 5.5223,
      "step": 2440
    },
    {
      "epoch": 65.26315789473684,
      "grad_norm": 25.390945434570312,
      "learning_rate": 1.1687500000000001e-05,
      "loss": 5.4784,
      "step": 2480
    },
    {
      "epoch": 66.3157894736842,
      "grad_norm": 81.70599365234375,
      "learning_rate": 1.0854166666666667e-05,
      "loss": 5.4446,
      "step": 2520
    },
    {
      "epoch": 67.36842105263158,
      "grad_norm": 30.602882385253906,
      "learning_rate": 1.0020833333333332e-05,
      "loss": 5.3098,
      "step": 2560
    },
    {
      "epoch": 68.42105263157895,
      "grad_norm": 23.538394927978516,
      "learning_rate": 9.187500000000001e-06,
      "loss": 5.4437,
      "step": 2600
    },
    {
      "epoch": 69.47368421052632,
      "grad_norm": 65.4513168334961,
      "learning_rate": 8.354166666666667e-06,
      "loss": 5.3217,
      "step": 2640
    },
    {
      "epoch": 70.52631578947368,
      "grad_norm": 21.81036376953125,
      "learning_rate": 7.5208333333333335e-06,
      "loss": 5.3911,
      "step": 2680
    },
    {
      "epoch": 71.57894736842105,
      "grad_norm": 21.587907791137695,
      "learning_rate": 6.687500000000001e-06,
      "loss": 5.3211,
      "step": 2720
    },
    {
      "epoch": 72.63157894736842,
      "grad_norm": 69.81565856933594,
      "learning_rate": 5.854166666666667e-06,
      "loss": 5.116,
      "step": 2760
    },
    {
      "epoch": 73.6842105263158,
      "grad_norm": 28.424556732177734,
      "learning_rate": 5.020833333333333e-06,
      "loss": 5.1371,
      "step": 2800
    },
    {
      "epoch": 74.73684210526316,
      "grad_norm": 14.789178848266602,
      "learning_rate": 4.1875e-06,
      "loss": 5.1651,
      "step": 2840
    },
    {
      "epoch": 75.78947368421052,
      "grad_norm": 17.214509963989258,
      "learning_rate": 3.354166666666667e-06,
      "loss": 5.0846,
      "step": 2880
    },
    {
      "epoch": 76.84210526315789,
      "grad_norm": 144.82855224609375,
      "learning_rate": 2.5208333333333335e-06,
      "loss": 5.0452,
      "step": 2920
    },
    {
      "epoch": 77.89473684210526,
      "grad_norm": 55.73275375366211,
      "learning_rate": 1.6875000000000001e-06,
      "loss": 4.9982,
      "step": 2960
    },
    {
      "epoch": 78.94736842105263,
      "grad_norm": 22.088788986206055,
      "learning_rate": 8.541666666666667e-07,
      "loss": 4.9426,
      "step": 3000
    },
    {
      "epoch": 80.0,
      "grad_norm": 12.161073684692383,
      "learning_rate": 2.0833333333333335e-08,
      "loss": 4.9881,
      "step": 3040
    }
  ],
  "logging_steps": 40,
  "max_steps": 3040,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 80,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9.260172472341234e+19,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}