Skip to content

Several minor tweak to vllm benchmark workflow #114

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.

Sign up for GitHub

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

Open
huydhn wants to merge 3 commits into
base: main
Choose a base branch
from
Open

Several minor tweak to vllm benchmark workflow #114

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
c4dcf9f
[no ci] Several minor tweak to vllm benchmark workflow
huydhn Dec 13, 2025
7ba6367
[no ci] Several minor tweak to vllm benchmark workflow
huydhn Dec 13, 2025
e5d0271
[no ci] Merge branch 'main' into vllm-benchmark-tweaks-1213
huydhn Dec 13, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 6 additions & 3 deletions .github/scripts/generate_vllm_benchmark_matrix.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,7 +19,7 @@
"linux.rocm.gpu.gfx942.1",
"linux.24xl.spr-metal",
"linux.24xl.gnr",
"linux.arm64.m7g.4xlarge",
# "linux.arm64.m7g.4xlarge", # TODO (huydhn): This is not working yet
"linux.dgx.b200",
"linux.hpu.gaudi3.8",
],
Expand Down Expand Up @@ -60,7 +60,7 @@
"linux.rocm.gpu.gfx942.8": "rocm",
"linux.24xl.spr-metal": "cpu",
"linux.24xl.gnr": "cpu",
"linux.arm64.m7g.4xlarge": "arm64-cpu",
# "linux.arm64.m7g.4xlarge": "arm64-cpu", # TODO (huydhn): This is not working yet
"linux.hpu.gaudi3.8": "hpu",
}

Expand Down Expand Up @@ -254,7 +254,10 @@ def generate_benchmark_matrix(
# Gather all possible benchmarks
for platform in sorted(platforms):
selected_models = []
for file in glob.glob(f"{benchmark_configs_dir}/{platform}/*.json"):
# Only need to parse serving config because it has all the models and
# their tensor_parallel_size field. The latter is used to find the runner
# with the right capacity
for file in glob.glob(f"{benchmark_configs_dir}/{platform}/*serving*.json"):
with open(file) as f:
try:
configs = json.load(f)
Expand Down
4 changes: 2 additions & 2 deletions .github/workflows/vllm-benchmark.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,8 +2,8 @@ name: vLLM Benchmark

on:
schedule:
# Run every 12 hours
- cron: '0 */12 * * *'
# Run daily at 1:15 AM PST
- cron: '15 9 * * *'
workflow_dispatch:
inputs:
vllm_branch:
Expand Down