Bio
Works at Hugging Face open-source team, author of the book Vision Language Models with Hugging Face published by O'Reilly.
Sessions (3)
Additional Speaker Details
Using publicly available information we constructed an analysis to help you get a feel for this speaker before deciding to attend their session.
Noyan is a Hugging Face developer advocacy engineer who turns the newest open-source multimodal and agentic tooling into things you can actually build with. Her recent AI Engineer Singapore talk live-demoed agents that fine-tune models autonomously (18 models for $11.40) using HF skills, MCP, and traces. Attend for a hands-on, current view of where open VLMs and self-training agents are headed from someone shipping the libraries and co-writing on them.
Speaking style
Demo-drivenDeeply technicalBattle-testedClear teacher
Recent talks (1)
GitHub
@merveenoyan- merveenoyan/smol-vision★ 1,950Recipes for shrinking, optimizing, customizing cutting edge vision models.
- merveenoyan/my_notes★ 1,653Personal notes/cheatsheets on data science, ML, and computer science.
- merveenoyan/siglip★ 310Projects based on SigLIP (Zhai et al., 2023) and Hugging Face transformers integration.
- merveenoyan/ml-roadmap★ 205A machine learning learning pathway / roadmap.
- merveenoyan/awesome-osml-for-devs★ 197List of resources and libraries for developers building with off-the-shelf open-source ML.
Recent writing (2)
- Vision Language Models (Better, faster, stronger) · blog · May 2025
- SmolVLM: Redefining small and efficient multimodal models · paper · Apr 2025
Podcasts & interviews (2)
- The Future of Vision in ML | Merve Noyan | HF Podcast #1 · Hugging Face Podcast · Mar 2026
- Are World Models the Next Big Thing? | Merve Noyan · Hugging Face Podcast · Apr 2026