Skip to content

A Hub GGUF this llama.cpp cannot load is flagged before download (0868, stacked on #19) - #21

Merged
lpalbou merged 3 commits into
mainfrom
feat/gguf-arch-before-download
Oct 29, 2026
Merged

lpalbou merged 3 commits into
mainfrom
feat/gguf-arch-before-download

Conversation

@lpalbou

@lpalbou lpalbou commented Oct 29, 2026

Copy link
Copy Markdown
Owner

Backlog 0868. Stacked on #19 (0867); retarget to main once #19 lands.

With --hub, _hub_model_facts keeps the Hub's gguf.architecture. _build_artifact checks it against libllama using 0867's gguf_unloadable_reason:

  • Cannot load: loadable: false, unloadable_reason (the sentence) and downloadable: false, plus a fit note. Never pre-selected.
  • Known: loadable: true.
  • Cannot be told: null, and the build stays offered (fail-safe).

Docs: docs/models.md. Test: test_a_hub_gguf_this_llamacpp_cannot_load_is_flagged_before_download (FakeHfApi now serves a gguf block). Mutation-checked. tests/config passes except 4 voice-probe tests that fail on the base branch too (environment).

…agged before download (Hub gguf.architecture; loadable false, sentence, not downloadable, never pre-selected); test; docs
@lpalbou
lpalbou changed the base branch from feat/gguf-architecture-support-check to main October 29, 2026 06:41
…-download

# Conflicts:
#	CHANGELOG.md
#	abstractcore/config/model_catalog.py
#	tests/config/test_model_catalog.py
…-download

# Conflicts:
#	tests/config/test_model_catalog.py
@lpalbou
lpalbou merged commit 5288d47 into main Oct 29, 2026
11 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant