10 questions · need 7/10 to pass.
Q1."Shrink a Model to a Quarter of Its Size (Quantization)" — which of these claims is supported by the module?
single
Q2.For "How Do You Know Your Fine-Tune Actually Worked?", which detail or constraint from the module is accurate?
single
Q3.Which fact about "Teach Preferences Without a Reward Model (DPO)" matches the mechanism the module covered?
single
Q4.When applying "Teach a Base Model to Follow Instructions (SFT)" in practice, which of these holds?
single
Q5.For "Which Fine-Tuning Method Fits Your GPU? (Full, LoRA, QLoRA)", which detail or constraint from the module is accurate?
single
Q6.Which statement about how "Fine-Tune Big Models on Small GPUs (LoRA)" actually works is correct?
single
Q7.Which definition of "Which Fine-Tuning Method Fits Your GPU? (Full, LoRA, QLoRA)" matches what the module established?
single
Q8.Which statement about how "Serve Your Model to Real Users (vLLM, TGI, Ollama)" actually works is correct?
single
Q9.When applying "How Chat Models Learn From Human Feedback (RLHF)" in practice, which of these holds?
single
Q10.Which of these correctly identifies the role of "Fine-Tune Big Models on Small GPUs (LoRA)" in the broader system?
single