Rooba
/
lightning
mirror of https://github.com/Lightning-AI/lightning.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
lightning/.azure/gpu-tests-pytorch.yml

216 lines
8.0 KiB
YAML
Raw Blame History

# Python package
# Create and test a Python package on multiple Python versions.
# Add steps that analyze code, save the dist with the build record, publish to a PyPI-compatible index, and more:
# https://docs.microsoft.com/azure/devops/pipelines/languages/python
trigger:
tags:
include:
- '*'
branches:
include:
- "master"
- "release/*"
- "refs/tags/*"
pr:
branches:
include:
- "master"
- "release/*"
paths:
include:
- ".actions/**"
- ".azure/gpu-tests-pytorch.yml"
- "examples/run_pl_examples.sh"
- "examples/pytorch/basics/backbone_image_classifier.py"
- "examples/pytorch/basics/autoencoder.py"
- "requirements/pytorch/**"
- "src/lightning/__about__.py"
- "src/lightning/__init__.py"
- "src/lightning/__main__.py"
- "src/lightning/__setup__.py"
- "src/lightning/__version__.py"
- "src/lightning/pytorch/**"
- "src/pytorch_lightning/*"
- "tests/tests_pytorch/**"
- "pyproject.toml" # includes pytest config
- "requirements/fabric/**"
- "src/lightning/fabric/**"
- "src/lightning_fabric/*"
exclude:
- "requirements/*/docs.txt"
- "*.md"
- "**/*.md"
jobs:
- job: testing
# how long to run the job before automatically cancelling
timeoutInMinutes: "80"
# how much time to give 'run always even if cancelled tasks' before stopping them
cancelTimeoutInMinutes: "2"
strategy:
matrix:
'PyTorch & strategies': # this uses torch 1.12 as not all strategies support 1.13 yet
image: "pytorchlightning/pytorch_lightning:base-cuda-py3.9-torch1.12-cuda11.6.1"
scope: "strategies"
PACKAGE_NAME: "pytorch"
'PyTorch | latest':
image: "pytorchlightning/pytorch_lightning:base-cuda-py3.9-torch1.13-cuda11.7.1"
scope: ""
PACKAGE_NAME: "pytorch"
'PyTorch | future':
image: "pytorchlightning/pytorch_lightning:base-cuda-py3.10-torch1.13-cuda11.7.1"
scope: "future"
PACKAGE_NAME: "pytorch"
'Lightning | latest':
image: "pytorchlightning/pytorch_lightning:base-cuda-py3.9-torch1.13-cuda11.7.1"
scope: ""
PACKAGE_NAME: "lightning"
pool: lit-rtx-3090
variables:
DEVICES: $( python -c 'print("$(Agent.Name)".split("_")[-1])' )
FREEZE_REQUIREMENTS: "1"
PIP_CACHE_DIR: "/var/tmp/pip"
container:
image: $(image)
# default shm size is 64m. Increase it to avoid:
# 'Error while creating shared memory: unhandled system error, NCCL version 2.7.8'
options: "--gpus=all --shm-size=2gb -v /var/tmp:/var/tmp"
workspace:
clean: all
steps:
- bash: |
echo "##vso[task.setvariable variable=CUDA_VISIBLE_DEVICES]$(DEVICES)"
cuda_ver=$(python -c "import torch ; print(''.join(map(str, torch.version.cuda.split('.')[:2])))")
echo "##vso[task.setvariable variable=CUDA_VERSION_MM]$cuda_ver"
echo "##vso[task.setvariable variable=TORCH_URL]https://download.pytorch.org/whl/cu${cuda_ver}/torch_stable.html"
scope=$( python -c 'n = "$(PACKAGE_NAME)" ; print(dict(pytorch="pytorch_lightning").get(n, n))' )
echo "##vso[task.setvariable variable=COVERAGE_SOURCE]$scope"
displayName: 'set env. vars'
- bash: |
echo $(DEVICES)
echo $CUDA_VISIBLE_DEVICES
echo $CUDA_VERSION_MM
echo $TORCH_URL
echo $COVERAGE_SOURCE
lspci | egrep 'VGA|3D'
whereis nvidia
nvidia-smi
which python && which pip
python --version
pip --version
pip list
displayName: 'Image info & NVIDIA'
- bash: |
PYTORCH_VERSION=$(python -c "import torch; print(torch.__version__.split('+')[0])")
for fpath in `ls requirements/**/*.txt`; do \
python ./requirements/pytorch/adjust-versions.py $fpath ${PYTORCH_VERSION}; \
done
condition: ne(variables['scope'], 'future')
displayName: 'Adjust dependencies'
- bash: pip install -e .[extra,test,examples] -U --find-links ${TORCH_URL}
displayName: 'Install package & dependencies'
- bash: |
pip install -U -q pip
pip uninstall -y torch torchvision
pip install "torch==2.0.0" "torchvision==0.15.0" -U --pre \
-f "https://download.pytorch.org/whl/test/cu${CUDA_VERSION_MM}/torch_test.html" \
-f "https://download.pytorch.org/whl/nightly/cu${CUDA_VERSION_MM}/torch_nightly.html"
python -c "from torch import __version__ as ver; assert ver.startswith('2.0.0'), ver"
condition: eq(variables['scope'], 'future')
displayName: 'bump to future'
- bash: pip uninstall -y -r requirements/pytorch/strategies.txt
condition: ne(variables['scope'], 'strategies')
displayName: 'Uninstall strategies'
- bash: |
set -e
pip install -r requirements/pytorch/strategies.txt --find-links ${TORCH_URL}
python requirements/pytorch/check-avail-strategies.py
condition: eq(variables['scope'], 'strategies')
displayName: 'Install strategies'
- bash: |
set -e
pip list
python requirements/collect_env_details.py
python -c "import torch ; mgpu = torch.cuda.device_count() ; assert mgpu == 2, f'GPU: {mgpu}'"
python requirements/pytorch/check-avail-extras.py
displayName: 'Env details'
- bash: bash .actions/pull_legacy_checkpoints.sh
displayName: 'Get legacy checkpoints'
- bash: python -m pytest pytorch_lightning
workingDirectory: src
condition: eq(variables['PACKAGE_NAME'], 'pytorch')
displayName: 'Testing: PyTorch doctests'
- bash: |
pip install -q -r .actions/requirements.txt
python .actions/assistant.py copy_replace_imports --source_dir="./tests/tests_pytorch" \
--source_import="lightning.fabric,lightning.pytorch" \
--target_import="lightning_fabric,pytorch_lightning"
python .actions/assistant.py copy_replace_imports --source_dir="./examples/pytorch/basics" \
--source_import="lightning.fabric,lightning.pytorch" \
--target_import="lightning_fabric,pytorch_lightning"
condition: eq(variables['PACKAGE_NAME'], 'pytorch')
displayName: 'Adjust tests & examples'
- bash: python -m coverage run --source ${COVERAGE_SOURCE} -m pytest --ignore benchmarks -v --durations=50
workingDirectory: tests/tests_pytorch
env:
PL_RUN_CUDA_TESTS: "1"
displayName: 'Testing: PyTorch standard'
timeoutInMinutes: "35"
- bash: bash run_standalone_tests.sh
workingDirectory: tests/tests_pytorch
env:
PL_USE_MOCKED_MNIST: "1"
PL_RUN_CUDA_TESTS: "1"
PL_STANDALONE_TESTS_SOURCE: $(COVERAGE_SOURCE)
displayName: 'Testing: PyTorch standalone tests'
timeoutInMinutes: "35"
- bash: bash run_standalone_tasks.sh
workingDirectory: tests/tests_pytorch
env:
PL_USE_MOCKED_MNIST: "1"
PL_RUN_CUDA_TESTS: "1"
displayName: 'Testing: PyTorch standalone tasks'
timeoutInMinutes: "10"
- bash: |
python -m coverage report
python -m coverage xml
python -m coverage html
python -m codecov --token=$(CODECOV_TOKEN) --commit=$(Build.SourceVersion) \
--flags=gpu,pytest,${COVERAGE_SOURCE} --name="GPU-coverage" --env=linux,azure
ls -l
workingDirectory: tests/tests_pytorch
displayName: 'Statistics'
- script: |
set -e
bash run_pl_examples.sh --trainer.accelerator=gpu --trainer.devices=1
bash run_pl_examples.sh --trainer.accelerator=gpu --trainer.devices=2 --trainer.strategy=ddp
bash run_pl_examples.sh --trainer.accelerator=gpu --trainer.devices=2 --trainer.strategy=ddp --trainer.precision=16
workingDirectory: examples
env:
PL_USE_MOCKED_MNIST: "1"
displayName: 'Testing: PyTorch examples'
- bash: python -m pytest benchmarks -v --maxfail=2 --durations=0
workingDirectory: tests/tests_pytorch
env:
PL_RUN_CUDA_TESTS: "1"
displayName: 'Testing: PyTorch benchmarks'
Reference in New Issue View Git Blame Copy Permalink