Skip to content

Support qmatmul with different dims tensors #17596

Support qmatmul with different dims tensors

Support qmatmul with different dims tensors #17596

Workflow file for this run

name: pull
on:
pull_request:
push:
branches:
- main
- release/*
workflow_dispatch:
concurrency:
group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.sha }}-${{ github.event_name == 'workflow_dispatch' }}-${{ github.event_name == 'schedule' }}
cancel-in-progress: true
jobs:
gather-models:
runs-on: ubuntu-22.04
outputs:
models: ${{ steps.gather-models.outputs.models }}
steps:
- uses: actions/checkout@v3
with:
submodules: 'false'
- uses: actions/setup-python@v4
with:
python-version: '3.10'
- name: Extract the list of models to test
id: gather-models
run: |
set -eux
PYTHONPATH="${PWD}" python .ci/scripts/gather_test_models.py --event "${GITHUB_EVENT_NAME}"
test-setup-linux-gcc:
name: test-setup-linux-gcc
uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
strategy:
matrix:
include:
- build-tool: cmake
fail-fast: false
with:
runner: linux.2xlarge
docker-image: executorch-ubuntu-22.04-gcc9
submodules: 'true'
ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
timeout: 90
script: |
# The generic Linux job chooses to use base env, not the one setup by the image
CONDA_ENV=$(conda env list --json | jq -r ".envs | .[-1]")
conda activate "${CONDA_ENV}"
BUILD_TOOL=${{ matrix.build-tool }}
PYTHON_EXECUTABLE=python bash .ci/scripts/setup-linux.sh "${BUILD_TOOL}"
# Build and test ExecuTorch with the add model on portable backend.
PYTHON_EXECUTABLE=python bash .ci/scripts/test.sh "add" "${BUILD_TOOL}" "portable"
test-models-linux:
name: test-models-linux
uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
needs: gather-models
strategy:
matrix: ${{ fromJSON(needs.gather-models.outputs.models) }}
fail-fast: false
with:
runner: ${{ matrix.runner }}
docker-image: executorch-ubuntu-22.04-clang12
submodules: 'true'
ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
timeout: ${{ matrix.timeout }}
script: |
# The generic Linux job chooses to use base env, not the one setup by the image
CONDA_ENV=$(conda env list --json | jq -r ".envs | .[-1]")
conda activate "${CONDA_ENV}"
MODEL_NAME=${{ matrix.model }}
BUILD_TOOL=${{ matrix.build-tool }}
BACKEND=${{ matrix.backend }}
DEMO_BACKEND_DELEGATION=${{ matrix.demo_backend_delegation }}
PYTHON_EXECUTABLE=python bash .ci/scripts/setup-linux.sh "${BUILD_TOOL}"
# Build and test ExecuTorch
PYTHON_EXECUTABLE=python bash .ci/scripts/test.sh "${MODEL_NAME}" "${BUILD_TOOL}" "${BACKEND}" "${DEMO_BACKEND_DELEGATION}"
test-llama-runner-linux:
name: test-llama-runner-linux
uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
strategy:
matrix:
dtype: [fp32]
build-tool: [buck2, cmake]
mode: [portable, xnnpack+custom, xnnpack+custom+qe]
fail-fast: false
with:
runner: linux.2xlarge
docker-image: executorch-ubuntu-22.04-clang12
submodules: 'true'
ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
timeout: 900
script: |
# The generic Linux job chooses to use base env, not the one setup by the image
CONDA_ENV=$(conda env list --json | jq -r ".envs | .[-1]")
conda activate "${CONDA_ENV}"
DTYPE=${{ matrix.dtype }}
BUILD_TOOL=${{ matrix.build-tool }}
MODE=${{ matrix.mode }}
# Setup executorch
PYTHON_EXECUTABLE=python bash .ci/scripts/setup-linux.sh buck2
# Install requirements for export_llama
PYTHON_EXECUTABLE=python bash examples/models/llama2/install_requirements.sh
# Test llama2
PYTHON_EXECUTABLE=python bash .ci/scripts/test_llama.sh stories110M.pt "${BUILD_TOOL}" "${DTYPE}" "${MODE}"
test-llama-runner-linux-android:
name: test-llama-runner-linux-android
uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
strategy:
matrix:
include:
- build-tool: cmake
fail-fast: false
with:
runner: linux.2xlarge
docker-image: executorch-ubuntu-22.04-clang12-android
submodules: 'true'
ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
timeout: 90
script: |
# The generic Linux job chooses to use base env, not the one setup by the image
CONDA_ENV=$(conda env list --json | jq -r ".envs | .[-1]")
conda activate "${CONDA_ENV}"
BUILD_TOOL=${{ matrix.build-tool }}
PYTHON_EXECUTABLE=python \
bash .ci/scripts/build_llama_android.sh "${BUILD_TOOL}"
test-custom-ops-linux:
name: test-custom-ops-linux
uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
strategy:
matrix:
include:
- build-tool: buck2
- build-tool: cmake
fail-fast: false
with:
runner: linux.2xlarge
docker-image: executorch-ubuntu-22.04-clang12
submodules: 'true'
ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
timeout: 90
script: |
# The generic Linux job chooses to use base env, not the one setup by the image
CONDA_ENV=$(conda env list --json | jq -r ".envs | .[-1]")
conda activate "${CONDA_ENV}"
BUILD_TOOL=${{ matrix.build-tool }}
PYTHON_EXECUTABLE=python bash .ci/scripts/setup-linux.sh "${BUILD_TOOL}"
# Test custom ops
PYTHON_EXECUTABLE=python bash examples/portable/custom_ops/test_custom_ops.sh "${BUILD_TOOL}"
test-selective-build-linux:
name: test-selective-build-linux
uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
strategy:
matrix:
include:
- build-tool: buck2
- build-tool: cmake
fail-fast: false
with:
runner: linux.2xlarge
docker-image: executorch-ubuntu-22.04-clang12
submodules: 'true'
ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
timeout: 90
script: |
# The generic Linux job chooses to use base env, not the one setup by the image
CONDA_ENV=$(conda env list --json | jq -r ".envs | .[-1]")
conda activate "${CONDA_ENV}"
BUILD_TOOL=${{ matrix.build-tool }}
PYTHON_EXECUTABLE=python bash .ci/scripts/setup-linux.sh "${BUILD_TOOL}"
# Test selective build
PYTHON_EXECUTABLE=python bash examples/selective_build/test_selective_build.sh "${BUILD_TOOL}"
test-export-llava-linux:
name: test-export-llava-linux
uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
strategy:
fail-fast: false
with:
runner: linux.12xlarge
docker-image: executorch-ubuntu-22.04-clang12
submodules: 'true'
ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
timeout: 90
script: |
# The generic Linux job chooses to use base env, not the one setup by the image
CONDA_ENV=$(conda env list --json | jq -r ".envs | .[-1]")
conda activate "${CONDA_ENV}"
PYTHON_EXECUTABLE=python bash .ci/scripts/setup-linux.sh "cmake"
# install Llava requirements
bash examples/models/llama2/install_requirements.sh
bash examples/models/llava/install_requirements.sh
# run export_llava.sh
python examples/models/llava/export_llava.py
# verify file exists
if [ ! -f "llava_combined_xnnpack.pte" ]; then
echo "llava_combined_xnnpack.pte not found!"
exit 1
fi
# run python unittest
python -m unittest examples.models.llava.test.test_llava
test-quantized-aot-lib-linux:
name: test-quantized-aot-lib-linux
uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
strategy:
matrix:
include:
- build-tool: cmake
fail-fast: false
with:
runner: linux.2xlarge
docker-image: executorch-ubuntu-22.04-clang12
submodules: 'true'
ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
timeout: 90
script: |
# The generic Linux job chooses to use base env, not the one setup by the image
CONDA_ENV=$(conda env list --json | jq -r ".envs | .[-1]")
conda activate "${CONDA_ENV}"
BUILD_TOOL=${{ matrix.build-tool }}
PYTHON_EXECUTABLE=python bash .ci/scripts/setup-linux.sh "${BUILD_TOOL}"
PYTHON_EXECUTABLE=python bash examples/xnnpack/quantization/test_quantize.sh "${BUILD_TOOL}" mv2
test-pybind-build-linux:
name: test-pybind-build-linux
uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
strategy:
matrix:
include:
- build-tool: cmake
fail-fast: false
with:
runner: linux.2xlarge
docker-image: executorch-ubuntu-22.04-clang12
submodules: 'true'
ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
timeout: 90
script: |
# The generic Linux job chooses to use base env, not the one setup by the image
CONDA_ENV=$(conda env list --json | jq -r ".envs | .[-1]")
conda activate "${CONDA_ENV}"
# build module for executorch.extension.pybindings.portable_lib
BUILD_TOOL=${{ matrix.build-tool }}
PYTHON_EXECUTABLE=python \
EXECUTORCH_BUILD_XNNPACK=ON \
EXECUTORCH_BUILD_PYBIND=ON \
bash .ci/scripts/setup-linux.sh "${BUILD_TOOL}"
# see if we can import the module successfully
python -c "from executorch.extension.pybindings import portable_lib; print('success!')"
test-binary-size-linux-gcc:
name: test-binary-size-linux-gcc
uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
strategy:
fail-fast: false
with:
runner: linux.2xlarge
docker-image: executorch-ubuntu-22.04-gcc9
submodules: 'true'
ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
timeout: 90
script: |
# The generic Linux job chooses to use base env, not the one setup by the image
CONDA_ENV=$(conda env list --json | jq -r ".envs | .[-1]")
conda activate "${CONDA_ENV}"
# build module for executorch.extension.pybindings.portable_lib
bash test/build_size_test.sh
strip cmake-out/test/size_test
output=$(ls -la cmake-out/test/size_test)
arr=($output)
size=${arr[4]}
# threshold=48120 on devserver with gcc11.4
# todo(lfq): update once binary size is below 50kb.
threshold="51504"
if [[ "$size" -le "$threshold" ]]; then
echo "Success $size <= $threshold"
else
echo "Fail $size > $threshold"
exit 1
fi
test-binary-size-linux:
name: test-binary-size-linux
uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
strategy:
fail-fast: false
with:
runner: linux.2xlarge
docker-image: executorch-ubuntu-22.04-clang12
submodules: 'true'
ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
timeout: 90
script: |
# The generic Linux job chooses to use base env, not the one setup by the image
CONDA_ENV=$(conda env list --json | jq -r ".envs | .[-1]")
conda activate "${CONDA_ENV}"
# build module for executorch.extension.pybindings.portable_lib
bash test/build_size_test.sh
strip cmake-out/test/size_test
output=$(ls -la cmake-out/test/size_test)
arr=($output)
size=${arr[4]}
# threshold=48120 on devserver with gcc11.4
# todo(lfq): update once binary size is below 50kb.
threshold="51768"
if [[ "$size" -le "$threshold" ]]; then
echo "Success $size <= $threshold"
else
echo "Fail $size > $threshold"
exit 1
fi
unittest:
uses: ./.github/workflows/_unittest.yml
with:
docker-image: executorch-ubuntu-22.04-clang12
unittest-arm:
uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
strategy:
matrix:
include:
- build-tool: buck2
with:
runner: linux.2xlarge
docker-image: executorch-ubuntu-22.04-arm-sdk
submodules: 'true'
ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
timeout: 90
script: |
set -eux
# The generic Linux job chooses to use base env, not the one setup by the image
CONDA_ENV=$(conda env list --json | jq -r ".envs | .[-1]")
conda activate "${CONDA_ENV}"
BUILD_TOOL=${{ matrix.build-tool }}
# Setup MacOS dependencies as there is no Docker support on MacOS atm
PYTHON_EXECUTABLE=python \
EXECUTORCH_BUILD_PYBIND=ON \
EXECUTORCH_BUILD_ARM_BAREMETAL=ON \
.ci/scripts/setup-linux.sh "${BUILD_TOOL}"
source .ci/scripts/utils.sh
# Install Arm dependencies
install_arm
# Run pytest with coverage
pytest -c /dev/null -v -n auto --cov=./ --cov-report=xml backends/arm/test