LoRA fine-tuning on your Mac (Project 4).

1. 1.Try #28 Reverse Linked List – untimed, AI-free

   Linked List · Easy. Aim for 15–20 minutes. No assistants, no autocomplete.

   Open on LeetCode (opens in a new tab)
2. 2.Stuck at 25 minutes? Use the NeetCode method

   Read or watch the explanation, close it, then re-code the solution from memory.

   NeetCode 150 explanations (opens in a new tab)
3. 3.Write the Big-O

   Add a comment with the time and space complexity, and why.

4. 4.Log it

   Mark it Solved or Revisit, with minutes and whether you solved it unaided.

   Log #28 on the Problems page
5. 5.10 min revision

   One item from reviews/relearn.md.

Companion video

Stanford CME295 Transformers & LLMs | Autumn 2025 | Lecture 5 – LLM tuningStanford Online · Mon–Wed

1. 1.Try #29 Merge Two Sorted Lists – untimed, AI-free

   Linked List · Easy. Aim for 15–20 minutes. No assistants, no autocomplete.

   Open on LeetCode (opens in a new tab)
2. 2.Stuck at 25 minutes? Use the NeetCode method

   Read or watch the explanation, close it, then re-code the solution from memory.

   NeetCode 150 explanations (opens in a new tab)
3. 3.Write the Big-O

   Add a comment with the time and space complexity, and why.

4. 4.Log it

   Mark it Solved or Revisit, with minutes and whether you solved it unaided.

   Log #29 on the Problems page
5. 5.10 min revision

   One item from reviews/relearn.md.

1. 1.Try #30 Linked List Cycle – untimed, AI-free

   Linked List · Easy. Aim for 15–20 minutes. No assistants, no autocomplete.

   Open on LeetCode (opens in a new tab)
2. 2.Stuck at 25 minutes? Use the NeetCode method

   Read or watch the explanation, close it, then re-code the solution from memory.

   NeetCode 150 explanations (opens in a new tab)
3. 3.Write the Big-O

   Add a comment with the time and space complexity, and why.

4. 4.Log it

   Mark it Solved or Revisit, with minutes and whether you solved it unaided.

   Log #30 on the Problems page
5. 5.10 min revision

   One item from reviews/relearn.md.

1. 1.Work through: mlx-lm – LoRA guide (LORA.md)

   ≈30 min

   Open (opens in a new tab)
2. 2.Work through: Hugging Face LLM Course – Chapter 11

   ≈45 min

   Open (opens in a new tab)
3. 3.Do

   Read; plan your dataset.

4. 4.Check you’re done

   Done when: p4/PLAN.md committed.

Companion video

QLoRA—How to Fine-tune an LLM on a Single GPU (w/ Python Code)Shaw Talebi · Thu

1. 1.Do

   Generate ≈150 synthetic adverts, label them with P1, hand-check 30; split train/valid/test.

2. 2.Check you’re done

   Done when: Three JSONL files.

1. 1.Catch up first

   Finish anything unticked from Mon–Fri this week before stretch work.

   Week 16 sessions
2. 2.Then stretch work

   Hand-check 30 more examples.

Companion video

EASIEST Way to Fine-Tune a LLM and Use It With OllamaTech With Tim · Flex

1. 1.Start with 20 min revision

   Explain what attention does.

2. 2.Do

   uv add "mlx-lm[train]"; mlx_lm.lora … --train --iters 300 --batch-size 2; base vs fine-tuned on the test set (JSON-valid rate, exact match per field).

3. 3.Check you’re done

   Done when: Before/after table committed.

4. 4.Commit

   projects/p4-mlx-lora/

   $ git add projects/p4-mlx-lora/

   $ git commit -m "Week 16: Train + evaluate"

   $ git push

1. 1.Do

   2 ablations; a model card (data, settings, results, limits, when not to fine-tune); weekly review.

2. 2.Check you’re done

   Done when: Model card committed.

3. 3.Commit

   P4 MODEL_CARD.md

4. 4.Weekly review (last 30 min)

   3 wins, 3 confusions, 1 thing to re-learn – written at the bottom of this page and saved with your progress.

   Open the weekly review