JevK5: open decision weights with CUDA and GGUF runtimes

Distilled Qwen models answer Choice, Noul and Score questions by reading option-token logits. The project releases 4B and 9B weights, fitted temperatures, a TypeSafe-compatible server and runtimes for CUDA or GGUF.

Notes

The v0.3 4B model card reports 87/111 correct on JevBench's public hard tier, ECE 0.054 and p50 13.2 ms on easy and standard items with an H100 and CUDA graphs.

source-backed — Public repository, docs or live artifact. About this label

Destinations

Availability & demo checks

Report a problem →

Also linked

• JevK5 4B weights ↗
• JevK5 9B weights ↗

Filed under

• #local
• #training
• #calibration
• #runtime

In collections

• Benchmarks & independent tests →
• Local & replica models →