Courses & Learning

Build a Language Model on Your Mac

Flavio Copes' free 29-lesson course on training a small language model with Language Model Builder on a Mac, covering tokens and embeddings, designing the run, pre-training and reading loss, fine-tuning with SFT and DPO, and comparing checkpoints in a lab report.

Curriculum

6 modules · 29 lessons

1

Plan the experiment

Install the app, set realistic expectations, and define what you will measure.
1. 01 What we are building
2. 02 Install the app and check your Mac
3. 03 What a small model can and cannot do
4. 04 Start a model lab notebook
5. 05 Check your understanding: plan the experiment
2

See what the model sees

Inspect tokens, embeddings, attention, and the transformer path.
1. 06 Tokens and token IDs
2. 07 Run the tokenizer playground
3. 08 Explore embedding space
4. 09 Trace a transformer prediction
5. 10 Check your understanding: model foundations
3

Design the run

Choose the model shape, tokenizer, data, and practical resource budget.
1. 11 Choose a model blueprint
2. 12 Parameters, memory, and training estimates
3. 13 Choose a tokenizer
4. 14 Select and prepare a dataset
5. 15 Check your understanding: design the run
4

Pre-train and observe

Run training over time, read loss, sample checkpoints, and diagnose results.
1. 16 Start a pre-training run
2. 17 Pause and return later
3. 18 Read training and validation loss
4. 19 Checkpoints and sampling
5. 20 Diagnose poor results
6. 21 Check your understanding: pre-train and observe
5

Shape model behavior

Use SFT, preference data, chat, and inspection tools deliberately.
1. 22 Run supervised fine-tuning
2. 23 Preference data and DPO
3. 24 Chat with your model
4. 25 Inspect the model with visualization tools
5. 26 Check your understanding: shape behavior
6

Compare and report

Compare checkpoints fairly and document what the experiment proved.
1. 27 Compare checkpoints fairly
2. 28 Write the model lab report
3. 29 Final check: compare and report