This project explores the capabilities of Haystack, an open-source framework for building search systems. We dive into various components and use cases to understand its potential in creating powerful, AI-driven search solutions.
- Document retrieval
- Question answering
- Semantic search
- And more!
- Clone the repository:
git clone https://github.com/90barricade93/haystack-explore- Navigate to the project directory:
cd haystack-explore
Set up a virtual environment:
python3 -m venv haystack_env
source haystack_env/bin/activateInstall dependencies:
pip install -r requirements.txtSet up API keys:
Create a `.env` file in the project root and add your API keys for any services you plan to use (e.g., OpenAI, Elasticsearch).Run any of the Python scripts to explore different Haystack features:
python3 first_rag_pipeline.py
python3 embedding_metadata_for_improved_retrieval.py
python3 extractive_qa_pipeline.py
python3 metadata_filtering.py
python3 file_type_preprocessing_index_pipeline.py
python3 serializing_pipelines.pyContributions are welcome! Please feel free to submit a Pull Request.
If you encounter any issues, please check The haystack tutorial or open an issue.
This project is licensed under the MIT License - see the LICENSE file for details.
