Fix duplicate Civitai weight selection#257
Merged
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What changed
Why
Civitai model version 2884631 advertises an fp16 primary and fp32 alternative under the same filename. The worker planned and downloaded both, overwrote the first with the second, and inflated disk preflight from about 18.2 GiB to 50.5 GiB. This made the live recovery ingest fail on a 10 GiB CPU pod before any CAS publish.
Validation
uv run --no-sync pytest -q tests/convert/test_civitai_gguf_select.py -k 'not ingest_civitai_gguf_classification' tests/convert/test_download_skip.py— 20 passed, 1 deselected.git diff --check, anduv build --offlinepass.Existing baseline failures
Untouched
origin/masteralready has one GGUF fixture failure, five mypy errors, anduv sync --locked --extra devcurrently fails because the mutabletorch==2.13.0+cu130wheel no longer matches the hashes recorded inuv.lock. This PR does not touch those paths or dependency hashes.