### What this PR does / why we need it? Add benchmark workflows ### Does this PR introduce _any_ user-facing change? No ### How was this patch tested? Run locally --------- Signed-off-by: wangli <wangli858794774@gmail.com>
161 lines
5.4 KiB
YAML
161 lines
5.4 KiB
YAML
#
|
|
# Copyright (c) 2025 Huawei Technologies Co., Ltd. All Rights Reserved.
|
|
# This file is a part of the vllm-ascend project.
|
|
#
|
|
# Licensed under the Apache License, Version 2.0 (the "License");
|
|
# you may not use this file except in compliance with the License.
|
|
# You may obtain a copy of the License at
|
|
#
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
#
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
# See the License for the specific language governing permissions and
|
|
# limitations under the License.
|
|
#
|
|
|
|
name: 'run benchmarks main'
|
|
|
|
on:
|
|
schedule:
|
|
- cron: '00 16 * * *'
|
|
workflow_dispatch:
|
|
|
|
# after merged, secrets will be available
|
|
# pull_request:
|
|
# branches:
|
|
# - 'main'
|
|
# - '*-dev'
|
|
# paths:
|
|
# - '.github/workflows/nightly_benchmarks.yaml'
|
|
|
|
|
|
# Bash shells do not use ~/.profile or ~/.bashrc so these shells need to be explicitly
|
|
# declared as "shell: bash -el {0}" on steps that need to be properly activated.
|
|
# It's used to activate ascend-toolkit environment variables.
|
|
defaults:
|
|
run:
|
|
shell: bash -el {0}
|
|
|
|
jobs:
|
|
test:
|
|
name: run benchmarks main
|
|
runs-on: 'linux-arm64-npu-static-8'
|
|
strategy:
|
|
matrix:
|
|
include:
|
|
- vllm_branch: v0.9.0
|
|
vllm_ascend_branch: main
|
|
container:
|
|
image: m.daocloud.io/quay.io/ascend/cann:8.1.rc1-910b-ubuntu22.04-py3.10
|
|
volumes:
|
|
- /usr/local/dcmi:/usr/local/dcmi
|
|
- /usr/local/bin/npu-smi:/usr/local/bin/npu-smi
|
|
- /usr/local/Ascend/driver/:/usr/local/Ascend/driver/
|
|
# Use self-host cache speed up pip and model download
|
|
- /home/action/.cache:/github/home/.cache/
|
|
options: >-
|
|
--device /dev/davinci0
|
|
--device /dev/davinci1
|
|
--device /dev/davinci_manager
|
|
--device /dev/devmm_svm
|
|
--device /dev/hisi_hdc
|
|
env:
|
|
HF_ENDPOINT: https://hf-mirror.com
|
|
HF_TOKEN: ${{ secrets.HF_TOKEN }}
|
|
ES_OM_DOMAIN: ${{ secrets.ES_OM_DOMAIN }}
|
|
ES_OM_AUTHORIZATION: ${{ secrets.ES_OM_AUTHORIZATION }}
|
|
steps:
|
|
- name: Check npu and CANN info
|
|
run: |
|
|
npu-smi info
|
|
cat /usr/local/Ascend/ascend-toolkit/latest/"$(uname -i)"-linux/ascend_toolkit_install.info
|
|
|
|
- name: Config mirrors
|
|
run: |
|
|
pip config set global.index-url https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple
|
|
|
|
- name: Install system dependencies
|
|
run: |
|
|
apt-get update -y
|
|
apt-get -y install git jq wget curl lsof gcc g++ cmake libnuma-dev
|
|
|
|
- name: Config git
|
|
run: |
|
|
git config --global --add safe.directory "$GITHUB_WORKSPACE"
|
|
git config --global url."https://gh-proxy.test.osinfra.cn/https://github.com/".insteadOf https://github.com/
|
|
|
|
|
|
- name: Checkout vllm-project/vllm-ascend repo
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ matrix.vllm_ascend_branch }}
|
|
|
|
- name: Checkout vllm-project/vllm repo
|
|
uses: actions/checkout@v4
|
|
with:
|
|
repository: vllm-project/vllm
|
|
path: ./vllm-empty
|
|
ref: ${{ matrix.vllm_branch }}
|
|
|
|
- name: Install vllm-project/vllm from source
|
|
working-directory: ./vllm-empty
|
|
run: |
|
|
VLLM_TARGET_DEVICE=empty pip install -e .
|
|
|
|
- name: Install vllm-project/vllm-ascend
|
|
run: |
|
|
pip install -e .
|
|
pip install -r benchmarks/requirements-bench.txt
|
|
|
|
- name: Checkout cosdt/elastic-tool
|
|
uses: actions/checkout@v4
|
|
with:
|
|
repository: cosdt/elastic-tool
|
|
path: ./elastic_tool
|
|
ref: 0.1.0-dev
|
|
|
|
- name: Install elastic_tool
|
|
working-directory: ./elastic_tool
|
|
run: |
|
|
pip install -e .
|
|
|
|
- name: Collect pr info from vllm-project/vllm-ascend
|
|
run: |
|
|
# Only get the pull request which may influences performance
|
|
git log --pretty=format:"%H %s" -- '**/*.py' ':!docs/*' ':!tests/*' ':!examples/*' > commit_log.txt
|
|
escli check commit_log.txt
|
|
|
|
- name: Run benchmark iteration
|
|
run: |
|
|
while IFS= read -r line || [[ -n "$line" ]]; do
|
|
commit_id=${line%% *}
|
|
commit_title=${line#* }
|
|
commit_time=$(git show -s --format=%cd $commit_hash --date=iso-strict)
|
|
commit_time_no_tz=${commit_time::19}
|
|
|
|
git checkout $commit_id
|
|
pip install -e .
|
|
|
|
echo "------------------------"
|
|
echo "commit_id: $commit_id"
|
|
echo "commit_title: $commit_title"
|
|
echo "commit_time: $commit_time_no_tz"
|
|
echo "vllm branch: ${{ matrix.vllm_branch }}"
|
|
echo "vllm-ascend branch: ${{ matrix.vllm_ascend_branch }}"
|
|
echo "------------------------"
|
|
|
|
bash benchmarks/scripts/run-performance-benchmarks.sh
|
|
# send the result to es
|
|
if [[ "${{ github.event_name }}" != "pull request" ]]; then
|
|
escli add --vllm_branch ${{ matrix.vllm_branch }} \
|
|
--vllm_ascend_branch ${{ matrix.vllm_ascend_branch }} \
|
|
--commit_id $commit_id \
|
|
--commit_title "$commit_title" \
|
|
--created_at "$commit_time_no_tz" \
|
|
--res_dir ./benchmarks/results
|
|
rm -rf ./benchmarks/results
|
|
fi
|
|
done < commit_log.txt
|