Raw Metadata XML
<pkgmetadata>
<maintainer type="person">
<email>web@bennypowers.com</email>
<name>Benny Powers</name>
</maintainer>
<longdescription lang="en">
Colibri runs GLM-5.2 (744B-parameter MoE) on consumer hardware with
as little as 25GB RAM. Pure C engine with zero runtime dependencies,
streaming experts from disk with an LRU cache. Supports MLA attention,
MTP speculative decoding, and optional CUDA GPU offloading.
</longdescription>
<use>
<flag name="bench">Install dependencies for quality benchmarks
(MMLU, HellaSwag, ARC)</flag>
<flag name="cuda">Build with NVIDIA CUDA GPU offloading for
resident expert tensors</flag>
<flag name="python">Install the coli Python CLI for model management,
chat, serving, and weight conversion</flag>
<flag name="rocm">Build with AMD ROCm/HIP GPU offloading for
resident expert tensors</flag>
</use>
<upstream>
<remote-id type="github">JustVugg/colibri</remote-id>
</upstream>
</pkgmetadata>