Nightly 2026-07-13: GGUF end-to-end — one .gguf file is the whole model (Qwen2.5 + LFM2.5) - #5

Merged

physics515 merged 6 commits into

mainfrom

mummu-nightly-2026-07-13

Jul 13, 2026

Commits on Jul 13, 2026

chore(deps): cargo update — uuid 1.23.5, zmij 1.0.22 (pins respected)

physics515
and
claude
committed
ca980ec
feat(p3): GGUF -> running model — full dequant suite + qwen2::load_from_gguf

physics515
and
claude
committed
03724d1
feat(p3): tokenizer-from-GGUF metadata — one .gguf file is the whole model

physics515
and
claude
committed
f6aee6f
docs: research fold — CubeCL quant primitives (P9), GPTQ/AWQ=compressed-tensors convention (P3), llama.cpp LFM2.5 parser fix upstream (P7)

physics515
and
claude
committed
d194ef4
feat(p3): LFM2.5 GGUF import — the hybrid loads from one file too

physics515
and
claude
committed
2876af9
feat(p3): registry learns GGUF — single-file model specs install end-to-end

physics515
and
claude
committed
f557f60