Repository navigation
[ROCm][CI] Raise the MI355 DeepSeek-R1 GSM8K startup wait to 1800s - #59666
Merged
AndreasKaratzas merged 2 commits intoOct 1, 2026
Merged
Conversation
The MI355 DeepSeek-R1 TP and DP GSM8K jobs, added in vllm-project#58683, load the 641 GiB checkpoint from the NFS Hugging Face cache. At nightly start the shards load at about 7.5 s each, so startup takes about 21 minutes and exceeds the 1200 s wait: both jobs failed this way in AMD CI nightly builds 13873 and 13954, and passed on a later retry when loading took about 11 minutes. Raise startup_max_wait_seconds to 1800 for the two MI355 configs. The test stays within the 2400 s pytest timeout and the 65-minute job limit. Signed-off-by: aarushjain29 <aarushi.jain2@amd.com>
aarushjain29
requested review from
AndreasKaratzas,
mgoin and
vadiklyutiy
as code owners
October 1, 2026 18:55
![claude[bot]](https://avatars.githubusercontent.com/in/1236702?s=60&v=4){kind=small}
AndreasKaratzas
approved these changes
Oct 1, 2026
AndreasKaratzas
enabled auto-merge (squash)
October 1, 2026 19:59
Member
|
/ci run |
github-actions
Bot
added
the
ready
|
✅ Triggered Buildkite CI #92483 for commit |
michael-gpu
pushed a commit
to practical-inference/vllm-ampere-extended
that referenced
this pull request
Oct 20, 2026
…llm-project#59666) Signed-off-by: aarushjain29 <aarushi.jain2@amd.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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 MI355 DeepSeek-R1 TP and DP GSM8K jobs, added in #58683, load the 641 GiB checkpoint from the NFS Hugging Face cache.
The first pass over the 163 shards takes about 21 minutes. With the MTP draft-model load and engine initialization, startup is about 24 minutes (projected 1416 s TP, 1427 s DP), which exceeds the 1200 s wait. 1800 s leaves about 6 minutes of headroom.
Raise startup_max_wait_seconds to 1800 for the two MI355 configs. The test stays within the 2400 s pytest timeout and the 65-minute job limit.