SUMMARY MODEL:

Model Params Size: 60492288
Model Params Size Formatted: 60.49 M
Model Disk Size: 242030465
Model Disk Size Formatted: 242.03 MB


TRAINING AND VALIDATION RESULTS:

Training batch size: 4
Validation batch size: 8
Total expected epochs: 4
Total expected trainig steps: 15052
Total expected trainig steps 2: 15052
Total trained epochs: 4.0
Total trained steps: 15052
Elapsed time: 7849.268085718155 seconds
Elapsed time (formatted): 02:10:49
Total flos: 8148659183026176.0
Total flos (formatted): 8.148659e+15
Best epoch val_loss: 0.5480290651321411
Best model checkpoint: E:/000_Tesis/test_executions/pretrain_utg4java\checkpoint-15052



SUMMARY DATASETS:

Loaded Dataset:
DatasetDict({
    train: Dataset({
        features: ['text'],
        num_rows: 15052
    })
    valid: Dataset({
        features: ['text'],
        num_rows: 1881
    })
    test: Dataset({
        features: ['text'],
        num_rows: 1882
    })
})

Tokenized Dataset:
DatasetDict({
    train: Dataset({
        features: ['input_ids'],
        num_rows: 15052
    })
    valid: Dataset({
        features: ['input_ids'],
        num_rows: 1881
    })
    test: Dataset({
        features: ['input_ids'],
        num_rows: 1882
    })
})