Prithvi-Raj-Dixit/Large_Language_Model_LLMs

1What is a Large Language Model?2 3An LLM is a neural network designed to generate text,audios,videos etc. and respond to human like text.4These LLMs are trained on massive amount of data.5Large- They are called large because thay are trained on billions of parameters.6Language Model- These models do a wide range of NLP tasks: question answering,translation,sentiment analysis(to find tone of user i.e. positive or negative) etc.7 8Earlier LLM models could not generate human like text.9 10Architecture of LLMs?11 12LLMs are based on Transformers archirecture based on paper "Attention Is All You Need".13 14AI15 └── ML16       └── DL17             └── LLMs(GenAI)18 19 20 21AI- A technology to mimic human behavior. It is broadest term.22Ex- Eliza chatbot is AI not ML(Eliza is not learning it uses predefined rules)23 24ML- It involves neural network and also decision tree. 25 26DL- It only involves neural networks. It involves many layers.27Ex- Stock price prediction28 29LLMs- They are subset of DL which involves text or tokens.30 31AI + ML + DL + LLM = GenAI32 33 34 35 36