Name the three stages of building a model and what each one teaches.
LLM & AILevel: IntermediateFine-tuning & Alignment
Name the three stages of building a model and what each one teaches.
Hints
- Identify the stage: pretraining, adaptation, generation or evaluation.
- Name the inputs, constraints and a case where the approach fails.
Answer
There are three stages, in order. Pretraining has the model read a massive pile of raw text predicting the next token, so it absorbs language, facts, and reasoning, though all it learns to do is continue text. SFT (supervised fine-tuning) has it imitate curated pairs of an instruction and an ideal answer, so it gains the behaviour of actually answering, in the right format. RLHF then tunes it on human rankings so it matches what people prefer in tone, helpfulness, and safety.
Source and licence: Skillumen — Awesome AI Engineer Interview Questions · CC0-1.0