feat(workflow): prefetch segmentation model for offline GPU nodes - #31
Merged
Conversation
…odes
Compute nodes without internet made Cellpose fail at run time while
downloading its weights ("Network is unreachable"). Add a fetch_model
localrule that runs on the (networked) submit host and caches the model
into shared $HOME before any GPU job; the segment rule now depends on it.
Its own log file avoids racing the shared steps.log.
Also fix the pixi [project] -> [workspace] deprecation warning and
document the offline-cluster and empty-SLURM-log cases.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The bug (caught by the new per-tile logs)
Segment jobs died with
message: None/ empty SLURM log. The per-tile log revealed the real trace:scicore's GPU nodes have no internet, so Cellpose couldn't download the
nucleiweights at run time. (Not GPU, not OOM — those were earlier red herrings;sacctshowed MaxRSS ~1 GB, ExitCode 1.)Fix
fetch_modellocalrule — runs on the networked submit host, caches the model into shared$HOME/.cellposebefore any GPU job.segmentnow depends on itsmodel.readymarker. No-ops formethod: threshold.localrules:so it's never submitted to an offline node. Dedicatedlogs/fetch_model.log(avoids racing the sequentialsteps.log).[project]→[workspace]deprecation.Test
Local run 13/13;
fetch_modelruns first as a localrule,logs/fetch_model.logcaptures its output,segmentwaits onmodel.ready. pixi.toml parses under[workspace].🤖 Generated with Claude Code