chore: pin baked hf CLI to 1.16.4 to match framework (no PyPI fetch at boot)#23
Merged
Merged
Conversation
… at boot) The framework's checkpoint_db shells out to `uvx hf@1.16.4 download ...` at model-init time, but the image baked/prefetched with hf@1.13.0. That version mismatch means uvx fetches hf 1.16.4 from PyPI on every pod boot (a runtime network dependency), and the prefetched HF cache (1.13.0 layout) isn't hit by the 1.16.4 runtime -> weights re-download from HF too. Align all hf invocations to 1.16.4 so the baked tool is reused and cache layout matches runtime. This is the tool half of removing the per-boot network dependency; the weights half (auxiliary encoder/VAE/sound-tokenizer served from deep-storage via a mounted HF cache + HF_HUB_OFFLINE=1) is a per-deployment change, not baked here. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Johan-de-R
changed the title
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
Bump the baked/prefetch
hfCLI from1.13.0→1.16.4throughout the Cosmos worker Dockerfile.Why
The framework (
cosmos_framework/utils/checkpoint_db.py) invokesuvx hf@1.16.4 download …at model-init to resolve its auxiliary weights. The image bakedhf@1.13.0(uv tool install) and prefetched with it too. Two consequences at every pod boot:uvx hf@1.16.4doesn't find a baked 1.16.4 tool → fetches it from PyPI (a runtime network dependency).Confirmed on live prod pods (Nano + the Super test): both do
uvx hf@1.16.4 download Qwen3-VL-… / Wan2.2 VAE / sound_tokenizerunauthenticated from public HF at boot.Scope
This is the tool half. The weights half — serving the auxiliary encoder/VAE/sound-tokenizer from deep-storage via a mounted HF cache +
HF_HUB_OFFLINE=1— is a per-deployment change (Super first, then Nano), not baked into this shared image, so it doesn't bloat the image or change Nano's behavior until Nano is migrated.Effect
Baked tool is reused (no PyPI at boot). Safe for both models (framework calls 1.16.4 regardless). Needed for the fully-offline boot we're building toward.
🤖 Generated with Claude Code