Updates for latest vLLM distributed example on Frontier/Orion#12
Draft
t-ramz wants to merge 7 commits into
Draft
Updates for latest vLLM distributed example on Frontier/Orion#12t-ramz wants to merge 7 commits into
t-ramz wants to merge 7 commits into
Conversation
added 3 commits
May 21, 2026 17:28
…Add multiple prompts for controlled experiment.
…de_lustre.sbatch: add required env variable for gpt-oss-120b
…running either gemma or gpt-oss at submit time. start_head_lustre.sh: Add capability to run model passed by batch config. testprompt_lustre.py: Add model detection. README.md: Add latest instructions'
secondspass
requested changes
May 27, 2026
| ray status | ||
|
|
||
| vllm serve --chat-template "./chattemplate.jinja" --tensor-parallel-size 8 --pipeline-parallel-size $NNODES --distributed-executor-backend ray "./astrollama-2-7b-base_abstract" --host 0.0.0.0 --port 8000 | ||
| vllm serve --chat-template "./chattemplate.jinja" --tensor-parallel-size 8 --pipeline-parallel-size $NNODES --distributed-executor-backend ray "./$RUN_MODEL" --host 0.0.0.0 --port 8000 --gpu-memory-utilization 0.75 No newline at end of file |
Contributor
There was a problem hiding this comment.
I would recommend removing the --chat-template flag. Models come with their own chat template file now, and it is automatically used. Providing one of our own isn't necessary anymore.
|
|
||
| [!NOTE] | ||
| `gpt-oss-120b` requires an additional step from login nodes. | ||
| Please additionally run the following commands to fetch the vocab file: |
Contributor
There was a problem hiding this comment.
I would recommend adding something informing the user that if they don't do this they will encounter an error like openai_harmony.HarmonyError: error downloading or loading vocab file: failed to download or load vocab file when they submit the job.
Contributor
|
Security approved talking about using the models. I added a warning box indicating that model use is approved only for this example and not beyond that |
added 3 commits
July 6, 2026 11:38
…_examples into feature/vllm-update
…support model param. README.md: add instructions for usage with burst buffer.
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
2 participants
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.
No description provided.