BookCapsule is an AI-driven book synopsis tool that distills complex text into its most essential insights and primary takeaways. With BookCapsule, you can quickly unpack the core of a book and save precious time without compromising comprehension.

• Chapter-by-Chapter Summarization: BookCapsule provides detailed summaries for each chapter, letting you concentrate on specific areas of interest.
• Whole Book Synopsis: In instances where the book doesn't have chapter divisions, BookCapsule condenses the entire text into a comprehensive overview.
• Powered by Natural Language Processing (NLP): With cutting-edge NLP techniques at its core, BookCapsule intelligently processes the text, capturing the most pertinent and informative content.
• Sleek User Interface: BookCapsule features a clean and intuitive interface, making the summarization process smooth and user-friendly, regardless of technical proficiency.

BookCapsule exploits the T5-small pretrained model from HuggingFace Transformers to generate precise and readable summaries. The process unfolds as follows:

1. Chunking: The text of the book is segmented into chunks, either by chapter or as a whole.
2. Tokenization: The chunks are tokenized using the T5Tokenizer to be compatible with the T5 model.
3. Summary Generation: The tokenized text is input into the T5ForConditionalGeneration model, which outputs summary token IDs.
4. Decoding: The summary token IDs are decoded back into intelligible text using the T5Tokenizer's decode() function, resulting in the final summary.

1. Clone the repository: git clone https://github.com/DeanLJY/BookCapsule.git
2. Install the necessary dependencies: pip install -r requirements.txt
3. Launch the application: python3 app.py

For more detailed instructions and advanced usage, please refer to the source code.

We value contributions from the community! If you wish to contribute to BookCapsule, please feel free to submit a pull request or open an issue. Your feedback and support are greatly appreciated!

BookCapsule is distributed under the MIT License.