📖 How to Build and Fine-Tune a Small Language Model
A Step-by-Step Guide for Beginners, Researchers, and Non-Programmers
How to Build and Fine-Tune a Small Language Model
Welcome! Below are the Google Colab notebooks for each chapter.
© 2025–2030 J. Paul Liu. All rights reserved.
This material is provided exclusively for readers of How to Build and Fine-Tune a Small Language Model. Please do NOT redistribute these Colab notebooks.
Chapter 2 – Let's Build GPT from Scratch
Open Colab NotebookChapter 3 – Quick Start: Fine-Tune Your First Model
Open Colab NotebookChapter 4 – Dataset Preparation
Open Colab NotebookChapter 5 – Model Architecture & Configuration
Open Colab NotebookChapter 6 – Training Loop & Monitoring
Open Colab NotebookChapter 7 – Evaluation & Benchmarks
Open Colab NotebookChapter 8 – Pre-training Your Own Language Model from Scratch
Open Colab NotebookChapter 9 – Stage 2: Supervised Fine-Tuning (SFT)
Open Colab NotebookChapter 10 – Stage 3: Direct Preference Optimization (DPO)
Open Colab NotebookChapter 11 – Production Deployment
Open Colab NotebookChapter 12 – Complete Production Projects & Ethics
Open Colab NotebookChapter 13 – Appendices
Open Colab NotebookBefore You Begin
- Visit Google Colab and make sure you can access Colab.
- If a notebook opens as plain text instead of inside Colab, choose Open With → Google Colab.
- If you don't see the Google Colab app, click Connect more apps.
- If your browser doesn't show "Open with Google Colab", download the
.ipynbfile and upload it manually from within Google Colab.
