Improving pre trained large language models with retrieval-augmented generation (RAG)

RAG is a method that uses external knowledge to improve LLMs’ answers. It has many benefits, such as accuracy, transparency, and knowledge updates by using own data. Will explain the development, components, and evaluation of RAG models, and suggests future research directions. The session will explain fundamental concepts, inner workings, significance, applications, and future of RAG by using real examples.

Key Takeaways

• RAG is important because it addresses some of the challenges of LLMs, such as presenting false, out-of-date, generic, or inaccurate information.
• RAG improves the relevance, accuracy, and usefulness of LLM output in various contexts, without the need to retrain the model by using your own Data.
• An idea of what the practical use of RAG is and how it can be implemented in current applications with live example.
• Evaluation of retrieval-augmented generation
  

————————————————————————————————————————————————————

Bio

Nasrullah Iqbal | Senior Enterprise architect |International Council for the Exploration of the Sea (ICES) | Denmark

Nasrullah Iqbal is also known as AI veteran – AI Scientist,  currently working as Senior Enterprise Architect and Tech Lead, and is engaged with AITU(Artificial Intelligence in IT university) research group and delivering guest lectures. 

Nasrullah has been working on AI since 2004 and has a master’s degree in Artificial Intelligence from Malardalen University, Sweden. He is also recognized as an AI veteran in the field. Before that, he studied software engineering at Hajvery University, Lahore, Pakistan.

Nasrullah began his professional career as a hardcore developer in 2006 and worked as a consultant for Danish companies until 2010.

Since 2010, He has been working with the International Council for the Exploration of the Sea (ICES) as a systems analyst and has a lot of practical experience with AI. He has designed and implemented the ICES cloud strategy as a lead. Now, now he is leading the AI strategy for ICES.