Skip to content

Fine-Tuning Your Own Customized Typhoon Model

This page contains materials from the special training AI Bootcamp: LLM Fine Tuning and Deployment, co-hosted by the Typhoon team and Float16. The one-day program covers both theory and hands-on practice. We’ve recorded the training sessions and collected all materials so you can learn at your own pace—anytime, anywhere.

Note: The video recordings are spoken in Thai.


You should have basic knowledge in the following areas before starting:

  1. Python programming skills
  2. Foundational knowledge of deep learning concepts
  3. Familiarity with a Linux or Unix-like environment

Topics covered in the first session:

  1. Intro to Typhoon LLMs and Fine Tuning
  2. Preparing for Fine Tuning
  3. Choosing the Right Fine Tuning Method
  4. Intro to LLM Evaluation
  5. Workshop Time 💻

Tip: click “Watch on YouTube” for the best viewing experience

👉 Download slides

1. Intro to Typhoon LLMs and Fine Tuning

2. Preparing for Fine Tuning

3. Choosing The Right Fine Tuning Method

4. Intro to LLM Evaluation

5. Workshop Time 💻

Fine-Tune Task: Sentiment Analysis The demonstration workshop fine-tunes a model for Sentiment Analysis tasks.

(For the morning session, use the file named “Fine-Tuning”)


Topics covered in the second session:

  1. Introduction to GGUF and Model Conversion
  2. Preparing the Model for GGUF Format
  3. Merging LoRA into the Full Model
  4. OpenAI API Compatibility (Endpoint Creation)
  5. Structured Outputs with Open-Source Models
  6. Workshop Time 💻

Tip: click “Watch on YouTube” for the best viewing experience

👉 Download slides

(For the afternoon session, use the file named “Deployment”)