C.W.K.
Stream
Quiz · 4 questions

🎯 Fine-Tuning — LoRA, QLoRA, DoRA, and Shipping the Result

Adapter-based fine-tuning end to end

Level 0Curious
0 XP0/51 lessons0/15 achievements
0/100 XP to next level100 XP to go0% complete

Quiz

01Why is full fine-tuning of a 70B model considered impractical on consumer Apple Silicon, and why does LoRA solve the problem?
Hint
It's about gradients and optimizer state, not just the weights themselves.
02Lesson 2 says the single biggest predictor of fine-tuning quality is what?
Hint
What's the thing the rest of the lesson hammers on?
03How do you turn on QLoRA in mlx-lm?
Hint
mlx-lm exposes DoRA via --fine-tune-type, but QLoRA isn't a separate setting.
04When should you fuse a trained adapter into the base model?
Hint
Iteration vs deployment is the key distinction.
Spotted a bug or have feedback on this page?Report an Issue

Comments 0

🔔 Reply notifications (sign in)
Sign inPlease sign in to comment.

No comments yet — be the first.