Mohit Mittal mittal-mohit

## deeplearningcourse
course link - https://www.coursera.org/learn/retrieval-augmented-generation-rag/lecture/Asarx/applications-of-rag

if something was included in training data of model, no additional info is required in the prompt for RAG

RAG is needed when we want to pair an LLM with info it wouldn't have had access to during training. retriver component of RAG will find useful info and add it to the prompt being sent to LLM so LLM doesnt have to find relevant info

RAG usage -
  1. code generation - generating specific code for a specific project requires specialised info
  2. company chatbots for policies, product SOPs
  3. AI assisted web search (web browsers earlier showed a list of page, now they show a relevant summary as well - this is RAG on internet)
	course link - https://www.coursera.org/learn/retrieval-augmented-generation-rag/lecture/Asarx/applications-of-rag

	if something was included in training data of model, no additional info is required in the prompt for RAG

	RAG is needed when we want to pair an LLM with info it wouldn't have had access to during training. retriver component of RAG will find useful info and add it to the prompt being sent to LLM so LLM doesnt have to find relevant info

	RAG usage -
	1. code generation - generating specific code for a specific project requires specialised info
	2. company chatbots for policies, product SOPs
	3. AI assisted web search (web browsers earlier showed a list of page, now they show a relevant summary as well - this is RAG on internet)