tev1

Data recipe and training code that fine-tune Qwen3.5-4B into an open-weight decision model on Together AI.

From the readme

tev1-4B-experimental We fine-tuned Qwen3.5-4B on Together AI to make decisions: give it context, a question, and 2–24 options, and it returns one answer letter. We're calling it tev1-4B-experimental. This repo contains the data recipe, training example, and saved results so you can fine-tune your own decision model. - Learn how to train your own classifier …

Details

Section: Open models and replicas
owner: togethercomputer
stars: 262
forks: 33
pushed: 2026-10-22
Language: Python
License: MIT

What it is

Form: Dataset or benchmark
Host agent: Standalone
Audience: App developers

Maturity

Docs

reports measured numbersmocked or unclear Jev call

Useful for

Model routing0.80

Classification and extraction0.77

Benchmarks and evals0.75

Best intent matches

classify support tickets or messages0.80

compare Jev with an LLM0.77

benchmark or calibrate Jev0.76

label data or build a dataset0.75

run Jev on open models without the vendor0.71

+ Submit a project ★ Star on GitHub