~ / open-implementations / trained-replicas / tev1
Tev1
Together AI's Jev-inspired experiment: Qwen3.5 fine-tunes that take state + question + 2–24 options and return one letter — openly next-token, not a non-autoregressive runtime. MIT, with full data recipe and a "train your own for $17" writeup. A 4B experimental build is also served on OpenRouter.
- section
- Open Implementations
- group
- Trained replicas
- source
- github.com
- repo
- togethercomputer/tev1
- stars
- 231 ▕██████░░░░░░▏
- 7-day gain
- +23
- forks
- 29
- language
- Python
- license
- MIT
- last pushed
- sep 24
- added to list
- sep 26
## history
2026-09-200 → 231 stars2026-10-13