Train your own Jev-like classifier for $17 with Qwen3.5 4B fine-tuning
nutlope · x · 2026-09-24
Together AI launched a tutorial and an open experimental model, together/Tev1-4B-experimental, fine-tuned on top of Qwen3.5 4B to replicate the Jev-style classifier.
- Jev-type models take a state plus predefined questions and quickly return a score, boolean, or multiple-choice answer — fast and extremely cheap to run
- The blog walks through the full fine-tuning process at a total cost of about $17
- Use cases include automated e-commerce return evaluation, ML paper categorization, and text sentiment scoring
Related event: Together Open-Sources Jev-Style Classifier, Fine-Tunable for $17(2 posts)→
More from Models
- Miles Brundage: Opus 5.5 Max 'works fairly hard', first time he skips defaulting to Ultracode — Miles_Brundage · 2026-09-24
- After one hour talking to Opus 5.5: "the industry is cooked" — TAbrodi · 2026-09-24
- Claude Opus 5.5 generates Backrooms found footage; reasoning effort changes the style — goodside · 2026-09-24
- Power user: Claude Opus 5.5 leads coding so hard his weekly usage barely moves — CtrlAltDwayne · 2026-09-24
- Instinct says leaked-doc complaint was hallucination, not a data breach — mon__lim · 2026-09-24
- Ex-OpenAI policy chief Miles Brundage: Opus 5.5 is indeed very good — Miles_Brundage · 2026-09-24