Apps & data pipelines

Lumma-Fev decision models

FrontiersMind's Lumma-Fev family comes with a Python client; its 27B model card describes multilingual training from Qwen3.8 for yes/no and choice answers.

Bookmark Keep this in your collection.

Leave a noteWhat would you try with this?

Only you can see your notes.

Source screenshot of Lumma-Fev decision models — SOURCE SCREENSHOTFull screenshot ↗

What it does

The maker reports multilingual training and comparisons with other decision models, without independent reproduction. The 27B card specifies yes/no and choice training, an 8,192-token context and substantial GPU memory needs. Lumma-Fev is a separate Apache-2.0 family, not Jev weights.

How you can use it

Read the card for the exact size you plan to run. Training for yes/no and choice is documented for 27B, not proof of every capability across the family. Check the memory needs and try cases with known answers before relying on the maker's comparisons.