checkpoint-1272/trainer_state.json

{
  "best_metric": 0.7450565099716187,
  "best_model_checkpoint": "/content/drive/MyDrive/Colab Notebooks/bert-base-multilingual-cased-finetuned-yiddish-experiment-3/checkpoint-1272",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1272,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.4716981132075472,
      "grad_norm": 59.220130920410156,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 12.7288,
      "step": 100
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 5.840357303619385,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 2.3607,
      "step": 200
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.0444412231445312,
      "eval_runtime": 1.4447,
      "eval_samples_per_second": 146.047,
      "eval_steps_per_second": 18.688,
      "step": 212
    },
    {
      "epoch": 1.4150943396226414,
      "grad_norm": 4.266133785247803,
      "learning_rate": 5e-06,
      "loss": 0.9508,
      "step": 300
    },
    {
      "epoch": 1.8867924528301887,
      "grad_norm": 5.5207600593566895,
      "learning_rate": 4.577702702702703e-06,
      "loss": 0.8155,
      "step": 400
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.8413578867912292,
      "eval_runtime": 1.5393,
      "eval_samples_per_second": 137.077,
      "eval_steps_per_second": 17.541,
      "step": 424
    },
    {
      "epoch": 2.358490566037736,
      "grad_norm": 3.453446626663208,
      "learning_rate": 4.155405405405405e-06,
      "loss": 0.7593,
      "step": 500
    },
    {
      "epoch": 2.830188679245283,
      "grad_norm": 4.256667613983154,
      "learning_rate": 3.733108108108108e-06,
      "loss": 0.7368,
      "step": 600
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.7711948752403259,
      "eval_runtime": 1.5544,
      "eval_samples_per_second": 135.739,
      "eval_steps_per_second": 17.369,
      "step": 636
    },
    {
      "epoch": 3.30188679245283,
      "grad_norm": 3.4198522567749023,
      "learning_rate": 3.310810810810811e-06,
      "loss": 0.7175,
      "step": 700
    },
    {
      "epoch": 3.7735849056603774,
      "grad_norm": 3.9418258666992188,
      "learning_rate": 2.888513513513514e-06,
      "loss": 0.7147,
      "step": 800
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.7569031715393066,
      "eval_runtime": 1.5508,
      "eval_samples_per_second": 136.058,
      "eval_steps_per_second": 17.41,
      "step": 848
    },
    {
      "epoch": 4.245283018867925,
      "grad_norm": 3.400341749191284,
      "learning_rate": 2.4662162162162163e-06,
      "loss": 0.6938,
      "step": 900
    },
    {
      "epoch": 4.716981132075472,
      "grad_norm": 4.095581531524658,
      "learning_rate": 2.043918918918919e-06,
      "loss": 0.692,
      "step": 1000
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.746055543422699,
      "eval_runtime": 1.5614,
      "eval_samples_per_second": 135.133,
      "eval_steps_per_second": 17.292,
      "step": 1060
    },
    {
      "epoch": 5.188679245283019,
      "grad_norm": 2.51070499420166,
      "learning_rate": 1.6216216216216219e-06,
      "loss": 0.684,
      "step": 1100
    },
    {
      "epoch": 5.660377358490566,
      "grad_norm": 3.3419506549835205,
      "learning_rate": 1.1993243243243243e-06,
      "loss": 0.6712,
      "step": 1200
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.7450565099716187,
      "eval_runtime": 1.5545,
      "eval_samples_per_second": 135.733,
      "eval_steps_per_second": 17.369,
      "step": 1272
    }
  ],
  "logging_steps": 100,
  "max_steps": 1484,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 668707919308800.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}