Automatic Spelling Correction

This project was inspired by a repository found at: Spelling Correction by oliverguhr.

Usage Guide:

1. Dataset Generation:

• Prepare a dataset suitable for your task in a directory named data as a TXT file.
• Run two scripts for dataset preparation:
  • convert_leipzig_data.sh
  • generate_dataset.py

2. Fine-Tuning:

• Utilize the fine-tuning script for training your model.
  • Refer to fine-tuning_script.py.

3. Environment Setup:

• All steps can be executed within a Colab notebook.
  • Refer to spelling_correction.ipynb.

4. Model Deployment:

• The final models have been uploaded to the Hugging Face Model Hub here.
• For deployment, consider using the provided Flask application:
  • Refer to app.py.

Testing Deployment:

• Test deployment using the provided link: Spelling Correction Deployment.