Here's the description for the fine-tuning session:

[SEA AI WEEK] Fine-Tuning Your Own LLM from an OSS Model

Want a model that knows your domain, speaks in your product's voice and runs on infrastructure you control? In this talk, you'll learn how teams take open-source LLMs and fine-tune them into custom models, and what it takes to do it well.

The talk walks through the full journey: when fine-tuning is (and isn't) the right choice compared with prompting or RAG, how to build a high-quality training dataset, parameter-efficient methods like LoRA and QLoRA, and how to tell whether your fine-tuned model actually performs better. You'll leave with a clear mental model and a practical playbook for building on open-source foundations.

Who it's for

  • Software engineers and ML/AI engineers

  • Technical founders and product leaders who are building AI products

  • Anyone weighing open-source models against off-the-shelf APIs

Key takeaways

  • How to choose the right open-source base model for your use case

  • What makes training data good (and what quietly ruins it)

  • Fine-tuning approaches and their trade-offs in cost, speed and quality

  • How to evaluate, deploy and keep improving a custom model

Details📅 Tuesday, October 27 · 1:00 – 2:30 PM📍 Thinkspace Seattle, 1700 Westlake Ave N #200, Seattle🎟 Spots are limited, so please register to save your seat

Speaker will be announced soon, so stay tuned.

AI Days at Thinkspace in partnership with Skillsheet. Details to Follow. For speaker, hosting, sponsorship opportunities contact: [email protected], [email protected]

Please note:

  1. By participating in the event, you agree to being in pictures or videos that may be published on our website, our partner's website or social media. 2. Our sponsors may message you about their offerings before or after the event. By RSVPing or attending the event, you agree to being contacted. You can opt out at any time.