-
-
Couldn't load subscription status.
- Fork 10.9k
Raise error for data-parallel with benchmark_throughput #16737
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Raise error for data-parallel with benchmark_throughput #16737
Conversation
|
👋 Hi! Thank you for contributing to the vLLM project. 💬 Join our developer Slack at https://slack.vllm.ai to discuss your PR in #pr-reviews, coordinate on features in #feat- channels, or join special interest groups in #sig- channels. Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging. To run CI, PR reviewers can either: Add 🚀 |
96c8576 to
e9240ba
Compare
Signed-off-by: Kartik Ramesh <kartikx2000@gmail.com>
Co-authored-by: Simon Mo <simon.mo@hey.com> Signed-off-by: Kartik Ramesh <kartikx2000@gmail.com>
Signed-off-by: Kartik Ramesh <kartikx2000@gmail.com>
e9240ba to
694e48a
Compare
|
@simon-mo @youkaichao Could I please get an auto-merge on this PR? Thank you. |
…#16737) Signed-off-by: Kartik Ramesh <kartikx2000@gmail.com> Co-authored-by: Simon Mo <simon.mo@hey.com> Signed-off-by: Yang Wang <elainewy@meta.com>
…#16737) Signed-off-by: Kartik Ramesh <kartikx2000@gmail.com> Co-authored-by: Simon Mo <simon.mo@hey.com> Signed-off-by: Frieda (Jingying) Huang <jingyingfhuang@gmail.com>
|
I encountered a problem when running this script, |
|
@Potabk What version of vLLM are you using? |
The latest benchmark script |
…#16737) Signed-off-by: Kartik Ramesh <kartikx2000@gmail.com> Co-authored-by: Simon Mo <simon.mo@hey.com>
|
I was referring to the version of vLLM you have installed. You can check this via |
Ok, never mind it, I think it is the inappropriate version, the latest version of vllm is ok |
…#16737) Signed-off-by: Kartik Ramesh <kartikx2000@gmail.com> Co-authored-by: Simon Mo <simon.mo@hey.com>
…#16737) Signed-off-by: Kartik Ramesh <kartikx2000@gmail.com> Co-authored-by: Simon Mo <simon.mo@hey.com> Signed-off-by: Agata Dobrzyniewicz <adobrzyniewicz@habana.ai>
…#16737) Signed-off-by: Kartik Ramesh <kartikx2000@gmail.com> Co-authored-by: Simon Mo <simon.mo@hey.com> Signed-off-by: Mu Huai <tianbowen.tbw@antgroup.com>
Running benchmark_throughput with data parallel results in a hang. This PR raises an error if
--data-parallel-sizeis passed in.Based on discussions in #16222
This is my first PR to vLLM so please let me know if I should add any tests. Thanks!