Jeff – Jev-compatible 0.8B decision models, trained at home, ~30 ms

Fine-tunes of Qwen3.5 and Gemma 4 for zero-shot classification - firelex/jeff
570 points by firelex 23 days ago github.com 223 comments

Summary

Firelex has released fine-tuned models, Jeff, based on Qwen3.5 and Gemma 4, designed for efficient zero-shot classification with a decision-making speed of around 30 milliseconds. This development could enhance real-time applications in various AI tasks.

Comments (223 of ) ⌛

Loading more comments... ⌛
Scroll down to load more comments
No comments available.
Type to search HN and engineering blog articles
Press Escape to close  |  ↑↓ to navigate  |  Enter to open