Clip / Other

4B param phone model already beats launch GPT-4

From Local Models: Trust, Control, Optimization — Carter Abdallah, NVIDIA · ≈40:58

Concrete capability benchmark showing how fast small local models have caught up to what was frontier just a couple years ago, underscoring the pace of on-device model progress.

What’s in it

Concrete capability benchmark showing how fast small local models have caught up to what was frontier just a couple years ago, underscoring the pace of on-device model progress.

Watch this moment ↗

Clip transcript

device on-device compute. >> You can run a 4 billion parameter model on your on your phone right now that is way more useful than GPT-4 was when it came out. And I think it's important for us to continue to to to focus on how do we best utilize that in the most meaningful way possible as well as chase

Intel