GAIC-VGG16 crop scoring on the AMD NPU
An IRON export of https://github.com/bo-zhang-cs/GAIC-Pytorch for AMD Ryzen AI NPUs: the compiled NPU kernels (.xclbin + instruction streams) and the packed weights the taconite-gaic Rust runtime replays, over XRT or directly over the amdxdna driver.
The kernels are compiled for NPU2 (AIE2P: Strix Point, Strix Halo, Krackan) and will not load on NPU1 (Phoenix, Hawk Point).
Download
hf download brishen/iron-gaic-vgg16-npu2 --local-dir gaic
# or, from an IRON checkout:
python scripts/hf_models.py download gaic --repo brishen/iron-gaic-vgg16-npu2 --out gaic
Usage
cargo install taconite-gaic # over XRT
# or, with no XRT at all (straight to the amdxdna driver):
cargo install taconite-gaic --no-default-features --features cli,direct
gaic gaic check # against the bundle's reference image
gaic gaic crop --out crops/ photo.jpg # best crop overall and at 1:1, 4:3, 16:9
See taconite-gaic (docs, source) for the full API. The bundle is exported by IRON's iron/applications/gaic.
Provenance
- Upstream model: https://github.com/bo-zhang-cs/GAIC-Pytorch (license:
mit; its terms apply to these weights) - IRON commit:
be4bb37 - Uploaded: 2026-09-23
- Files: 21, 29.2 MB