decider-0.8b for the Ryzen AI NPU

Mapika/decider-0.8b, a Qwen3.5-0.8B classifier, converted to run on the Ryzen AI NPU of AMD Strix Halo inside halogen-flash-server, beside the Flash model on the GPU and behind the same port. Questions go to the Discord. Bugs go to the server repository's issues.

These files run only inside the halogen-flash-server container, on the NPU. They will not load in transformers or any other runtime.

What is here

devices/             684 MB  the NPU program for this model, and its NOTICE.md
decider-0.8b.hnpw   2.30 GB  the weights, converted for the NPU
tokenizer/            20 MB  the tokenizer

Use it

Start halogen-flash-server 0.16.0 or later with HALOGEN_DOWNLOAD on and HALOGEN_NPU_MODELS=decider-0.8b. The server fetches these files into /models/npu/decider-0.8b/ and checks each one's size and checksum before it starts. Since 0.18.2 it takes texts up to 8,192 tokens, and with HALOGEN_NPU_DEC_RESUME=1 a request with several questions reads its text once. Give it a text and a question with 2 to 10 options, and it returns a probability for every option from one pass: a guard, a router, a classifier beside the Flash model.

The files at the head of this repository need halogen-flash-server 0.18.12 or later; each server version downloads the revision it was released with, so an older server keeps working unchanged.

The host needs the NPU driver, XRT with its NPU plugin, and the GPU's fabric clock held while the NPU works beside the GPU. The setup, every request shape, and how to run your own fine-tune of this model are in docs/NPU.md.

Source and license

Converted from Mapika/decider-0.8b at revision a0a01d6f8135298f400a8c856b355793012ae971, itself a fine-tune of Qwen/Qwen3.5-0.8B-Base. Apache-2.0, as the source. devices/NOTICE.md lists the third-party notices the NPU program carries.