{
  "best_metric": 0.19640253484249115,
  "best_model_checkpoint": "/kaggle/working/hubert-amharic/checkpoint-1500",
  "epoch": 12.121212121212121,
  "eval_steps": 500,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.40404040404040403,
      "grad_norm": 2.795124053955078,
      "learning_rate": 9.730094466936572e-06,
      "loss": 1.5155,
      "step": 100
    },
    {
      "epoch": 0.8080808080808081,
      "grad_norm": 2.900432825088501,
      "learning_rate": 9.460188933873145e-06,
      "loss": 1.175,
      "step": 200
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 3.376126766204834,
      "learning_rate": 9.195681511470986e-06,
      "loss": 0.7957,
      "step": 300
    },
    {
      "epoch": 1.6161616161616161,
      "grad_norm": 6.262876987457275,
      "learning_rate": 8.925775978407558e-06,
      "loss": 0.6227,
      "step": 400
    },
    {
      "epoch": 2.0202020202020203,
      "grad_norm": 13.498932838439941,
      "learning_rate": 8.65587044534413e-06,
      "loss": 0.5372,
      "step": 500
    },
    {
      "epoch": 2.0202020202020203,
      "eval_accuracy": 0.8929293155670166,
      "eval_loss": 0.3763836622238159,
      "eval_runtime": 22.9508,
      "eval_samples_per_second": 21.568,
      "eval_steps_per_second": 5.403,
      "step": 500
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 7.660800457000732,
      "learning_rate": 8.388663967611337e-06,
      "loss": 0.4195,
      "step": 600
    },
    {
      "epoch": 2.8282828282828283,
      "grad_norm": 10.333969116210938,
      "learning_rate": 8.118758434547908e-06,
      "loss": 0.4207,
      "step": 700
    },
    {
      "epoch": 3.2323232323232323,
      "grad_norm": 25.161924362182617,
      "learning_rate": 7.848852901484481e-06,
      "loss": 0.3231,
      "step": 800
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.32242974638938904,
      "learning_rate": 7.578947368421054e-06,
      "loss": 0.236,
      "step": 900
    },
    {
      "epoch": 4.040404040404041,
      "grad_norm": 4.83636474609375,
      "learning_rate": 7.309041835357625e-06,
      "loss": 0.2254,
      "step": 1000
    },
    {
      "epoch": 4.040404040404041,
      "eval_accuracy": 0.9111111164093018,
      "eval_loss": 0.3798251748085022,
      "eval_runtime": 23.0284,
      "eval_samples_per_second": 21.495,
      "eval_steps_per_second": 5.385,
      "step": 1000
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 10.075101852416992,
      "learning_rate": 7.039136302294197e-06,
      "loss": 0.2038,
      "step": 1100
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 5.713657379150391,
      "learning_rate": 6.76923076923077e-06,
      "loss": 0.2233,
      "step": 1200
    },
    {
      "epoch": 5.252525252525253,
      "grad_norm": 64.90503692626953,
      "learning_rate": 6.499325236167342e-06,
      "loss": 0.2077,
      "step": 1300
    },
    {
      "epoch": 5.656565656565657,
      "grad_norm": 33.007598876953125,
      "learning_rate": 6.229419703103914e-06,
      "loss": 0.1811,
      "step": 1400
    },
    {
      "epoch": 6.0606060606060606,
      "grad_norm": 2.3471884727478027,
      "learning_rate": 5.959514170040487e-06,
      "loss": 0.1699,
      "step": 1500
    },
    {
      "epoch": 6.0606060606060606,
      "eval_accuracy": 0.9535353779792786,
      "eval_loss": 0.19640253484249115,
      "eval_runtime": 22.503,
      "eval_samples_per_second": 21.997,
      "eval_steps_per_second": 5.51,
      "step": 1500
    },
    {
      "epoch": 6.4646464646464645,
      "grad_norm": 46.472137451171875,
      "learning_rate": 5.692307692307692e-06,
      "loss": 0.1676,
      "step": 1600
    },
    {
      "epoch": 6.8686868686868685,
      "grad_norm": 0.10037334263324738,
      "learning_rate": 5.4224021592442655e-06,
      "loss": 0.1468,
      "step": 1700
    },
    {
      "epoch": 7.2727272727272725,
      "grad_norm": 2.2828967571258545,
      "learning_rate": 5.152496626180837e-06,
      "loss": 0.1122,
      "step": 1800
    },
    {
      "epoch": 7.6767676767676765,
      "grad_norm": 2.2225501537323,
      "learning_rate": 4.882591093117409e-06,
      "loss": 0.1162,
      "step": 1900
    },
    {
      "epoch": 8.080808080808081,
      "grad_norm": 3.6428568363189697,
      "learning_rate": 4.6126855600539814e-06,
      "loss": 0.1245,
      "step": 2000
    },
    {
      "epoch": 8.080808080808081,
      "eval_accuracy": 0.9595959782600403,
      "eval_loss": 0.22902674973011017,
      "eval_runtime": 22.8105,
      "eval_samples_per_second": 21.701,
      "eval_steps_per_second": 5.436,
      "step": 2000
    },
    {
      "epoch": 8.484848484848484,
      "grad_norm": 0.09281215816736221,
      "learning_rate": 4.342780026990554e-06,
      "loss": 0.1027,
      "step": 2100
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 30.872848510742188,
      "learning_rate": 4.072874493927126e-06,
      "loss": 0.105,
      "step": 2200
    },
    {
      "epoch": 9.292929292929292,
      "grad_norm": 0.3525031507015228,
      "learning_rate": 3.8029689608636982e-06,
      "loss": 0.1184,
      "step": 2300
    },
    {
      "epoch": 9.696969696969697,
      "grad_norm": 0.40138208866119385,
      "learning_rate": 3.53306342780027e-06,
      "loss": 0.1048,
      "step": 2400
    },
    {
      "epoch": 10.1010101010101,
      "grad_norm": 0.025619197636842728,
      "learning_rate": 3.2631578947368423e-06,
      "loss": 0.0597,
      "step": 2500
    },
    {
      "epoch": 10.1010101010101,
      "eval_accuracy": 0.9636363387107849,
      "eval_loss": 0.22432678937911987,
      "eval_runtime": 22.7183,
      "eval_samples_per_second": 21.789,
      "eval_steps_per_second": 5.458,
      "step": 2500
    },
    {
      "epoch": 10.505050505050505,
      "grad_norm": 0.028187109157443047,
      "learning_rate": 2.9932523616734146e-06,
      "loss": 0.0525,
      "step": 2600
    },
    {
      "epoch": 10.909090909090908,
      "grad_norm": 0.020364606752991676,
      "learning_rate": 2.723346828609987e-06,
      "loss": 0.1105,
      "step": 2700
    },
    {
      "epoch": 11.313131313131313,
      "grad_norm": 18.78951644897461,
      "learning_rate": 2.453441295546559e-06,
      "loss": 0.078,
      "step": 2800
    },
    {
      "epoch": 11.717171717171716,
      "grad_norm": 48.03165817260742,
      "learning_rate": 2.183535762483131e-06,
      "loss": 0.1166,
      "step": 2900
    },
    {
      "epoch": 12.121212121212121,
      "grad_norm": 0.21182887256145477,
      "learning_rate": 1.913630229419703e-06,
      "loss": 0.0816,
      "step": 3000
    },
    {
      "epoch": 12.121212121212121,
      "eval_accuracy": 0.9555555582046509,
      "eval_loss": 0.27169349789619446,
      "eval_runtime": 22.618,
      "eval_samples_per_second": 21.885,
      "eval_steps_per_second": 5.482,
      "step": 3000
    },
    {
      "epoch": 12.121212121212121,
      "step": 3000,
      "total_flos": 7.685447856522912e+17,
      "train_loss": 0.2951181084314982,
      "train_runtime": 1630.5884,
      "train_samples_per_second": 18.205,
      "train_steps_per_second": 2.272
    }
  ],
  "logging_steps": 100,
  "max_steps": 3705,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.685447856522912e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}