Skip to content

chore(model gallery): 🤖 add 1 new models via gallery agent - #11841

Open
localai-org-maint-bot wants to merge 1 commit into
mudler:masterfrom
ci-forks:create-pull-request/patch-1788356630
Open

chore(model gallery): 🤖 add 1 new models via gallery agent#11841
localai-org-maint-bot wants to merge 1 commit into
mudler:masterfrom
ci-forks:create-pull-request/patch-1788356630

Conversation

@localai-org-maint-bot

Copy link
Copy Markdown
Collaborator

This PR was automatically created by the gallery agent workflow.

Summary:

  • Search Term: GGUF
  • Models Found: 15
  • Models Added: 1
  • Quantization: Q4_K_M
  • Processing Time: 470.418061ms

Added Models:

Bot commands

Maintainers (owner / member / collaborator) can control this PR
by leaving a comment with one of:

  • /gallery-agent recreate — close this PR; the next scheduled
    run will propose this model again (useful if the entry needs
    to be regenerated with fresh metadata).
  • /gallery-agent blacklist — close this PR and permanently
    prevent the gallery agent from ever reproposing this model.

Plain "Close" (without a command) is treated as a no-op: the
model may be reproposed by a future run.

Workflow Details:

  • Triggered by: schedule
  • Run ID: 33630073650
  • Commit: 7c921aa02093b98ae2e0368e4c1b3abd25670759

@localai-org-maint-bot

Copy link
Copy Markdown
Collaborator Author

Review pass. Blocking for the same reason as #11832, plus a name collision.

Placeholder checksums from a gated repo. Both files carry the literal mask the HF API returns for gated repos:

sha256: '****************************************************************'

orcarouter/Qwen3.8-27B-Uncensored-GGUF is gated: auto and returns HTTP 401 anonymously, so the download fails for users and the checksum could not verify anyway.

Name collision. This PR and #11847 both want the gallery name qwen3.8-27b-uncensored, from two different upstream repos. Only one can land under that name. #11847 points at JonathanColetti/Qwen3.8-27B-Uncensored-GGUF, which is ungated and whose checksums verify, so that is the better candidate for the name.

Suggest closing this one in favour of #11847 unless there is a reason to prefer the orcarouter build.

@localai-org-maint-bot

Copy link
Copy Markdown
Collaborator Author

Holding this one: the sha256 fields are not digests.

Every entry in this PR carries 64 literal asterisks as its checksum:

sha256: '****************************************************************'

I checked the raw bytes to rule out display masking — the file really does contain * characters.

That is worse than omitting the field. DownloadFileWithContext skips verification only when the digest is empty; a non-empty value means it verifies, compares the computed hex against the asterisks, deletes the download and returns SHA mismatch for file .... So the entry can never install, after pulling a Q4_K_M plus its mmproj.

Two other things worth a look while this is open:

  • The description is the upstream model card pasted verbatim, including GitHub > [!Note] / > [!Tip] callouts that will not render in the gallery UI, Qwen Cloud's hosted-API marketing, and a literal ... truncation marker.
  • The weights point at orcarouter/Qwen3.8-27B-Uncensored-GGUF — a vendor-hosted repackage rather than the publisher's own repo. Same vendor as docs(cloud-proxy): add OrcaRouter passthrough example #11657. Worth deciding deliberately whether vendor rehosts belong in the gallery.

Real digests can be read from HuggingFace's tree API (lfs.oid is the content sha256).

Signed-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
@localai-org-maint-bot
localai-org-maint-bot force-pushed the create-pull-request/patch-1788356630 branch from 81ba24a to 42031f1 Compare September 7, 2026 04:04
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.

2 participants