Free courses from frontend to fullstack and AI
Become an AI & ML Engineer with Cal Poly EPaCE — IBM-Certified Training
Overview
Google, IBM & Meta Certificates — All 10,000+ Courses at 40% Off
One annual plan covers every course and certificate on Coursera. 40% off for a limited time.
Get Full Access
Discover how to build a compact vision language model that outperforms much larger competitors from major tech companies in this 19-minute conference talk. Learn about the unexpected journey behind creating Moondream, an open-source vision model that achieves remarkable performance despite its small size, beating models four times larger and competing with offerings from Meta, Google, and OpenAI. Explore the technical challenges faced during development and understand why small models represent the future of AI technology. Gain insights into the democratization of AI through efficient model design that enables advanced computer vision capabilities on resource-constrained devices without sacrificing performance. Follow along as the speaker shares the story of accidental innovation and demonstrates how thinking small can lead to significant breakthroughs in making AI technology accessible to developers and businesses of all sizes.
Syllabus
Moondream: how does a tiny vision model slap so hard? — Vikhyat Korrapati
Taught by
AI Engineer