Skip to content

Benchmarks

Benchmarks #452

# This workflow periodically runs the benchmarks suite in benchmarks/
# using asv and publish the results, effectively updating
# the display website hosted in the pybamm-bench repo
# Steps:
# - Benchmark all commits since the last one that was benchmarked
# - Push results to pybamm-bench repo
# - Publish website
name: Benchmarks
on:
# Every day at 3 am UTC
schedule:
- cron: "0 3 * * *"
# Make it possible to trigger the
# workflow manually
workflow_dispatch:
env:
PYBAMM_DISABLE_TELEMETRY: "true"
jobs:
benchmarks:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
- name: Set up Python 3.12
uses: actions/setup-python@0b93645e9fea7318ecaed2b359559ac225c90a2b # v5.3.0
with:
python-version: 3.12
- name: Install Linux system dependencies
run: |
sudo apt-get update
sudo apt-get install gfortran gcc libopenblas-dev
- name: Set up uv
run: python -m pip install uv
- name: Install python dependencies
run: |
python -m uv pip install --upgrade pip wheel setuptools wget cmake casadi numpy
python -m uv pip install asv[virtualenv]
- name: Install SuiteSparse and SUNDIALS
run: python scripts/install_KLU_Sundials.py
- name: Run benchmarks
run: |
asv machine --machine "GitHubRunner"
asv run --machine "GitHubRunner" NEW --show-stderr -v
env:
SUNDIALS_INST: $HOME/.local
LD_LIBRARY_PATH: $HOME/.local/lib
- name: Upload results as artifact
uses: actions/upload-artifact@b4b15b8c7c6ac21ea08fcf65892d2ee8f75cf882 # v4.4.3
with:
name: asv_periodic_results
path: results
if-no-files-found: error
publish-results:
if: github.repository == 'pybamm-team/PyBaMM'
name: Push and publish results
needs: benchmarks
runs-on: ubuntu-latest
steps:
- name: Set up Python 3.12
uses: actions/setup-python@0b93645e9fea7318ecaed2b359559ac225c90a2b # v5.3.0
with:
python-version: 3.12
- name: Install asv
run: pip install asv
- name: Checkout pybamm-bench repo
uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
with:
repository: pybamm-team/pybamm-bench
token: ${{ secrets.BENCH_PAT }}
- name: Download results artifact(s)
uses: actions/download-artifact@fa0a91b85d4f404e444e00e005971372dc801d16 # v4.1.8
with:
path: results
merge-multiple: true
- name: Copy new results and push to pybamm-bench repo
env:
PUSH_BENCH_EMAIL: ${{ secrets.PUSH_BENCH_EMAIL }}
PUSH_BENCH_NAME: ${{ secrets.PUSH_BENCH_NAME }}
run: |
git config --global user.email "$PUSH_BENCH_EMAIL"
git config --global user.name "$PUSH_BENCH_NAME"
git add results
git commit -am "Add new results"
git push
- name: Publish results
run: |
asv publish
git fetch origin gh-pages:gh-pages
asv gh-pages