Telco-RAG is a specialized Retrieval-Augmented Generation (RAG) framework designed to tackle the unique challenges presented by the telecommunications industry, particularly when working with complex and rapidly evolving 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: Tailored specifically to handle the intricacies of telecommunications standards.
- Enhanced Query Processing: Utilizes a dual-stage query enhancement and retrieval process to improve the accuracy and relevance of generated responses.
- Hyperparameter Optimization: Carefully tuned to deliver the best performance by optimizing chunk sizes, context length, and embedding models.
- NN Router: A neural network-based router that improves the efficiency and accuracy of document retrieval, significantly reducing RAM usage.
- Open-Source: Freely available for the community to use, adapt, and improve upon.
The video is presented at 1.5x speed.
To get started with Telco-RAG, you'll need to clone the repository and set up the environment:
git clone https://github.com/netop-team/telco-rag.git
cd <repository-directory>
Prerequisites Python 3.8+ Uvicorn Other dependencies listed in requirements.txt
Installation Install the necessary Python packages:
cd .\Telco-RAG_api\
pip install -r requirements.txt
Running the API To launch the API server, use the following command:
cd .\Telco-RAG_api\
uvicorn api.deploy_api:app --reload
License This project is licensed under the MIT License - see the LICENSE file for details.