MeetMind RAG

Overview

MeetMind RAG is a Streamlit-based application designed for Retrieval-Augmented Generation (RAG). It enables text processing, vector-based retrieval, similarity search, and audio transcription. The project leverages chromadb for vector storage, sentence-transformers for embeddings, TfidfVectorizer for text similarity, and FFmpeg-dependent libraries for transcription.

Features

• Text Retrieval: Efficient vector search using chromadb.
• Similarity Search: TF-IDF and cosine similarity for text analysis.
• Audio Transcription: Processes audio inputs (requires FFmpeg).
• Streamlit Interface: Web-based UI for user interaction.
• Extensible: Modular structure for adding features.

Prerequisites

• Python: 3.11–3.13 (tested on 3.13)
• Git: For cloning the repository
• Virtual Environment: Recommended
• FFmpeg: For audio transcription
• Streamlit Cloud: For deployment (optional)

System Dependencies

• SQLite: 3.35.0+ (local; patched for Streamlit Cloud)
  • Check: sqlite3 --version
• FFmpeg: Required for transcription
  • Ubuntu: sudo apt install ffmpeg
  • macOS: brew install ffmpeg
  • Windows: Install via ffmpeg.org or Chocolatey

Getting Started

1. Clone the Repository

git clone https://github.com/sudhakar-r08/MeetMind.git
cd MeetMind

2. Set Up Virtual Environment

python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate

3. Install Dependencies

Create requirements.txt:

streamlit
chromadb>=0.5.0
numpy
scikit-learn
sentence-transformers
pysqlite3-binary
speech_recognition  # Adjust if using a different transcription library
pydub              # Adjust as needed

Install:

pip install -r requirements.txt

Install FFmpeg locally (see System Dependencies).

4. Run Locally

streamlit run meetmind_rag.py

Access at http://localhost:8501.

Project Structure

MeetMind/
├── meetmind_rag.py             # Main application script
├── requirements.txt            # Python dependencies
├── packages.txt                # System dependencies (e.g., ffmpeg)
├── venv/                       # Virtual environment (gitignore)
└── README.md                   # This file

Usage

• Use the Streamlit UI to query text or upload audio for transcription.
• Customize meetmind_rag.py to modify retrieval, similarity, or transcription logic.

Deployment to Streamlit Cloud

1. Push your repository to GitHub.
2. Deploy via Streamlit Cloud.
3. SQLite Fix:
   • Add to requirements.txt: pysqlite3-binary
   • Patch meetmind_rag.py at the top:

     import sys
     import pysqlite3
     sys.modules['sqlite3'] = sys.modules.pop('pysqlite3')

4. FFmpeg Fix:
   • Create packages.txt:

     ffmpeg
     

5. ChromaDB Space Import Fix:
   • Use chromadb>=0.5.0 in requirements.txt.
   • Replace Space references with chromadb.Client and create_collection:

     import chromadb
     client = chromadb.Client()
     collection = client.create_collection(name="my_collection")

6. Redeploy and check logs.

Troubleshooting

SQLite Version Error

Error: Your system has an unsupported version of sqlite3...

• Local: Upgrade SQLite or install pysqlite3-binary.
• Streamlit Cloud: Apply the SQLite patch above.
• Docs: ChromaDB Troubleshooting

FFmpeg Error

Error: Transcription failed: [Errno 2] No such file or directory: 'ffmpeg'

• Local: Install FFmpeg.
• Streamlit Cloud: Add ffmpeg to packages.txt.
• Docs: FFmpeg Documentation

ChromaDB ImportError

Error: cannot import name 'Space' from 'chromadb.api.types'

• Remove Space imports; use chromadb.Client and create_collection.
• Update ChromaDB: pip install --upgrade chromadb.
• Docs: ChromaDB API

Contributing

1. Fork the repository.
2. Create a branch: git checkout -b feature-branch
3. Commit changes: git commit -m "Add feature"
4. Push: git push origin feature-branch
5. Open a Pull Request.

License

MIT (assumed; add LICENSE file if needed).

Contact

For issues or feedback, use GitHub Issues or contact sudhakar-r08.