This project demonstrates how to build a Retrieval-Augmented Generation (RAG) application using farmer-specific data. The data is extracted from a PDF containing crop variety information for sugarcane, turmeric, bamboo, cashew nuts, and more. The application leverages LangChain and LanceDB to create a customizable and extensible FarmerGPT solution.
- Customizable Prompts: Easily adapt prompts to suit specific queries and use cases.
- Memory Support: Incorporates memory capabilities to retain context during interactions.
- PDF Integration: Processes and retrieves data from the provided PDF file.
. Use the Colab Notebook:
- To try FarmerGPT directly without setup, use the provided Google Colab notebook:
- Open the notebook, follow the instructions, and run the cells to interact with the application.
- LangChain: Framework for building applications powered by large language models.
- LanceDB: Vector database used for efficient document retrieval.
- Assisting farmers with queries about crop varieties.
- Providing tailored advice based on specific crop information.
- Serving as a reference for agricultural experts and enthusiasts.
You can build upon this template to include:
- Additional crop varieties.
- New functionalities such as integration with IoT devices.
- Multi-language support for wider accessibility.