-
Notifications
You must be signed in to change notification settings - Fork 109
Configure PyTorch x vLLM nightly dashboard #7614
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
Conversation
Signed-off-by: Huy Do <[email protected]>
Signed-off-by: Huy Do <[email protected]>
Signed-off-by: Huy Do <[email protected]>
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
torchci/components/benchmark_v3/configs/teams/vllm/pytorch_x_vllm_config.ts
Show resolved
Hide resolved
| name: "PyTorch x vLLM Benchmark", | ||
| route: `/benchmark/v3/dashboard/${PYTORCH_X_VLLM_BENCHMARK_ID}`, | ||
| info: "PyTorch x vLLM nightly benchmark. Powered by [code](TODO) and [benchmark configs](https://github.com/pytorch/pytorch-integration-testing/tree/main/vllm-benchmarks/benchmarks)", | ||
| description: "Pytorch x vLLM nightly benchmark using vllm bench", |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@huydhn recommend add comment that it's generated based on pinned vllm commit
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
also, maybe put this under vllm section: "vLLM Benchmarks ", instead of pytorch,
maybe also update the description for the vllmx pinned pytorch description, so that we can tell the difference
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I put it under PyTorch mainly because the users of this dashboard will be PT2 compiler looking at vllm bench from PyTorch point of view. The public vLLM v1 dashboard is a separate entity with a different group of users, maybe we would deprecate this in the future if there is no interest from vLLM maintainers in keeping it.
Signed-off-by: Huy Do <[email protected]>
Signed-off-by: Huy Do <[email protected]>
Signed-off-by: Huy Do <[email protected]>
Signed-off-by: Huy Do <[email protected]>
A new dashboard for PyTorch x vLLM nightly benchmark after pytorch/pytorch#170466 lands
Preview
Benchmark results are showing up on https://torchci-git-pytorch-x-vllm-dashboard-fbopensource.vercel.app/benchmark/v3/dashboard/pytorch_x_vllm_benchmark?renderGroupId=main&time.start=2025-12-12T00%3A00%3A00.000Z&time.end=2025-12-19T23%3A59%3A59.999Z&lcommit.commit=7f2cac9b0c4cb1a0dc36f9c1daa7c6596c69b69a&lcommit.workflow_id=20362340534&lcommit.date=2025-12-19T23%3A00%3A00Z&lcommit.branch=setup-vllm-nightly-benchmark&rcommit.commit=7f2cac9b0c4cb1a0dc36f9c1daa7c6596c69b69a&rcommit.workflow_id=20362340534&rcommit.date=2025-12-19T23%3A00%3A00Z&rcommit.branch=setup-vllm-nightly-benchmark&lbranch=setup-vllm-nightly-benchmark&rbranch=setup-vllm-nightly-benchmark for the test run https://github.com/pytorch/pytorch/actions/runs/20362340534/job/58585533990
cc @zou3519