Telco-RAG is a specialized Retrieval-Augmented Generation (RAG) framework designed to address the unique challenges of the telecommunications industry, particularly in handling the complexity and rapid evolution of 3GPP documents.
- Bornea, A.-L., Ayed, F., De Domenico, A., Piovesan, N., & Maatouk, A. (2024). Telco-RAG: Navigating the Challenges of Retrieval-Augmented Language Models for Telecommunications. arXiv preprint arXiv:2404.15939. DOI | Read the paper
- Custom RAG Pipeline: Specifically tailored to handle the complexities of telecommunications standards.
- Enhanced Query Processing: Implements a dual-stage query enhancement and retrieval process, improving the accuracy and relevance of generated responses.
- Hyperparameter Optimization: Optimized for the best performance by fine-tuning chunk sizes, context length, and embedding models.
- NN Router: A neural network-based router that enhances document retrieval efficiency while significantly reducing RAM usage.
- Open-Source: Freely available for the community to use, adapt, and improve.
The video is presented at 1.5x speed.
To get started with Telco-RAG, clone the repository and set up the environment:
git clone https://github.com/netop-team/telco-rag.git
cd telco-rag- Python 3.11
- Node.js
- Git LFS
Other dependencies are listed in requirements.txt.
Install the necessary Python packages and download the 3GPP knowledge database:
cd ./Telco-RAG_api
pip install -r requirements.txt
python setup.pyTo run the full pipeline, use the following commands:
npm install concurrently
npm install
npm run devThis will open two terminals: one for the frontend and one for the Telco-RAG backend. You can access the frontend via your browser at http://localhost:3000/.
On your first connection, ensure to specify a valid OpenAI API key in the settings.
If you only want to run the API server, use this command:
cd ./Telco-RAG_api
uvicorn api.deploy_api:app --reloadThis project is licensed under the MIT License. See the LICENSE file for details.