Introduction

Question 1

Power of LLM

Answer 1

Use ChatGPT API as developer tool to quickly build software applications

Question 2

Types of LLM

Answer 2

Base LLM
Instruction tuned LLM

Question 3

Base LLM

Answer 3

It predicts the next word, based on text training data from the internet and other sources e.g. if you write a sentence, it will complete it while if you write a question it will complete it by adding more questions because this is what is has learned from the internet.

Question 4

Instruction Tuned LLM

Answer 4

It is trained to follow the instructions.
These LLMs are first trained on base LLMs and then further fine tuned with inputs and outputs that are instructions using a technique called Reinforcement Learning with Human Feedback (RLHF) to make the system better follow instructions that are helpful, honest and harmless.