Running a Question-Answering System on Ray Serve at Deepset

Explore the process of running a question-answering system on Ray Serve in this 31-minute talk from Anyscale. Delve into the key architectural components of question-answering systems, including data stores, indexing pipelines, and querying pipelines. Learn about Haystack, an open-source framework that connects multiple transformer state-of-the-art NLP models into a single pipeline. Discover how to deploy GPU-empowered inference using Ray, assemble NLP models into pipelines, run Hugging Face models on Ray Serve, deploy NLP model pipelines, and access persistent storage from code deployed on Ray Serve. Gain valuable insights into enhancing your question-answering systems and leveraging Ray Serve for improved performance and scalability.