Repository navigation
docs(vast): record the GPU runner check on the released 0.22.0 - #44
Merged
Merged
Conversation
The review fixes of #42 were rerun on a card with the published image: the root-side watchdog, the cap after a torch probe, the CatBoost share, raising runs per card mid-run, deleting a competition mid-run, the neighbour-burst retry, a class on the other card and a 25-submission load. Only the fail-closed stop stays checked with stand-ins.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Follow-up to #42. The review fixes were rerun on the lab VM (2 × RTX 4090) with the published colloq-vast:0.22.0: root-side watchdog stops (10.4 GB and 19.8 GB runs), the torch cap after a find_spec probe (held at 5.25 GiB of 6141 MB), CatBoost gpu_ram_part 0.229 at 4 per card, a per-card raise while a whole-card run holds the card (new runs waited), deleting a competition mid-run (run stopped, card freed), the neighbour-burst retry (victim rerun and scored), a kaggle-gpu class on the other card, and 25 simultaneous submissions on 2 cards × 4 (run 22–29 s, last verdict 104 s). Only the fail-closed stop on an unreadable card stays checked with stand-ins. docs-gpu and vast-image tests: 34/34.