#double-quantization

2 questions from Fine-tuning & PEFT

Timed test

2 questions · 2 min · score and explanations at the end

Q 1 / 2Score 0 / 0

Shuffle questions

Question 1: Single choice

mediumAI-checkedFine-tuning & PEFT

QLoRA fine-tunes LoRA adapters on top of a base model stored in 4-bit precision. What does its "double quantization" do?

It also quantizes the quantization constants, saving more memory1
It quantizes every weight twice to reach 1-bit storage2
It swaps weights back to 16 bits whenever the loss gets worse3
It quantizes the LoRA adapters as well as the frozen base model weights4

Keys: 1–9 select · Enter check / next