workbench-example-hybrid-rag

The NVIDIA workbench-example-hybrid-rag is a chat application designed for Retrieval Augmented Generation (RAG) that integrates with NVIDIA APIs, NIM containers, and Hugging Face TGI on NVIDIA GPUs. Its primary use case is to facilitate the testing of RAG features with customizable document embedding and versatile inference options, including local, cloud, and microservice deployments. Notable features include support for various model quantization levels, local inference capabilities, and the ability to utilize a range of downloadable models and microservices seamlessly.