Fix missing job action button in CI report (#17270)

* use matrix.machine_type

* fix job names used in job_link

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
This commit is contained in:
Yih-Dar
2022-05-17 08:31:06 +02:00
committed by GitHub
parent 5a9957358c
commit 1ac2b8fa7f
3 changed files with 62 additions and 106 deletions

View File

@@ -73,7 +73,7 @@ jobs:
echo "::set-output name=test_map::$test_map"
run_tests_single_gpu:
name: Model Tests on single GPU
name: Model tests
needs: setup
# `dummy` means there is no test to run
if: contains(fromJson(needs.setup.outputs.matrix), 'dummy') != true
@@ -81,8 +81,8 @@ jobs:
fail-fast: false
matrix:
folders: ${{ fromJson(needs.setup.outputs.matrix) }}
machines: [single-gpu]
runs-on: [self-hosted, docker-gpu, '${{ matrix.machines }}']
machine_type: [single-gpu]
runs-on: [self-hosted, docker-gpu, '${{ matrix.machine_type }}']
container:
image: huggingface/transformers-all-latest-gpu
options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
@@ -117,22 +117,22 @@ jobs:
- name: Run all non-slow selected tests on GPU
working-directory: /transformers
run: |
python3 -m pytest -n 2 --dist=loadfile -v --make-reports=${{ matrix.machines }}_tests_gpu_${{ matrix.folders }} ${{ fromJson(needs.setup.outputs.test_map)[matrix.folders] }}
python3 -m pytest -n 2 --dist=loadfile -v --make-reports=${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }} ${{ fromJson(needs.setup.outputs.test_map)[matrix.folders] }}
- name: Failure short reports
if: ${{ failure() }}
continue-on-error: true
run: cat /transformers/reports/${{ matrix.machines }}_tests_gpu_${{ matrix.folders }}/failures_short.txt
run: cat /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}/failures_short.txt
- name: Test suite reports artifacts
if: ${{ always() }}
uses: actions/upload-artifact@v2
with:
name: ${{ matrix.machines }}_run_all_tests_gpu_${{ env.matrix_folders }}_test_reports
path: /transformers/reports/${{ matrix.machines }}_tests_gpu_${{ matrix.folders }}
name: ${{ matrix.machine_type }}_run_all_tests_gpu_${{ env.matrix_folders }}_test_reports
path: /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}
run_tests_multi_gpu:
name: Model Tests on multi GPUs
name: Model tests
needs: setup
# `dummy` means there is no test to run
if: contains(fromJson(needs.setup.outputs.matrix), 'dummy') != true
@@ -140,8 +140,8 @@ jobs:
fail-fast: false
matrix:
folders: ${{ fromJson(needs.setup.outputs.matrix) }}
machines: [multi-gpu]
runs-on: [self-hosted, docker-gpu, '${{ matrix.machines }}']
machine_type: [multi-gpu]
runs-on: [self-hosted, docker-gpu, '${{ matrix.machine_type }}']
container:
image: huggingface/transformers-all-latest-gpu
options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
@@ -178,29 +178,29 @@ jobs:
MKL_SERVICE_FORCE_INTEL: 1
working-directory: /transformers
run: |
python3 -m pytest -n 2 --dist=loadfile -v --make-reports=${{ matrix.machines }}_tests_gpu_${{ matrix.folders }} ${{ fromJson(needs.setup.outputs.test_map)[matrix.folders] }}
python3 -m pytest -n 2 --dist=loadfile -v --make-reports=${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }} ${{ fromJson(needs.setup.outputs.test_map)[matrix.folders] }}
- name: Failure short reports
if: ${{ failure() }}
continue-on-error: true
run: cat /transformers/reports/${{ matrix.machines }}_tests_gpu_${{ matrix.folders }}/failures_short.txt
run: cat /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}/failures_short.txt
- name: Test suite reports artifacts
if: ${{ always() }}
uses: actions/upload-artifact@v2
with:
name: ${{ matrix.machines }}_run_all_tests_gpu_${{ env.matrix_folders }}_test_reports
path: /transformers/reports/${{ matrix.machines }}_tests_gpu_${{ matrix.folders }}
name: ${{ matrix.machine_type }}_run_all_tests_gpu_${{ env.matrix_folders }}_test_reports
path: /transformers/reports/${{ matrix.machine_type }}_tests_gpu_${{ matrix.folders }}
run_tests_torch_cuda_extensions_single_gpu:
name: Torch CUDA extension tests on single GPU
name: Torch CUDA extension tests
needs: setup
if: contains(fromJson(needs.setup.outputs.matrix), 'deepspeed') || contains(fromJson(needs.setup.outputs.matrix), 'extended')
strategy:
fail-fast: false
matrix:
machines: [single-gpu]
runs-on: [self-hosted, docker-gpu, '${{ matrix.machines }}']
machine_type: [single-gpu]
runs-on: [self-hosted, docker-gpu, '${{ matrix.machine_type }}']
container:
image: nvcr.io/nvidia/pytorch:21.03-py3
options: --gpus 0 --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
@@ -227,29 +227,29 @@ jobs:
- name: Run all non-slow selected tests on GPU
# TODO: Here we pass all tests in the 2 folders for simplicity. It's better to pass only the identified tests.
run: |
python -m pytest -n 1 --dist=loadfile -v --make-reports=${{ matrix.machines }}_tests_torch_cuda_extensions_gpu tests/deepspeed tests/extended
python -m pytest -n 1 --dist=loadfile -v --make-reports=${{ matrix.machine_type }}_tests_torch_cuda_extensions_gpu tests/deepspeed tests/extended
- name: Failure short reports
if: ${{ failure() }}
continue-on-error: true
run: cat reports/${{ matrix.machines }}_tests_torch_cuda_extensions_gpu/failures_short.txt
run: cat reports/${{ matrix.machine_type }}_tests_torch_cuda_extensions_gpu/failures_short.txt
- name: Test suite reports artifacts
if: ${{ always() }}
uses: actions/upload-artifact@v2
with:
name: ${{ matrix.machines }}_run_tests_torch_cuda_extensions_gpu_test_reports
path: reports/${{ matrix.machines }}_tests_torch_cuda_extensions_gpu
name: ${{ matrix.machine_type }}_run_tests_torch_cuda_extensions_gpu_test_reports
path: reports/${{ matrix.machine_type }}_tests_torch_cuda_extensions_gpu
run_tests_torch_cuda_extensions_multi_gpu:
name: Torch CUDA extension tests on multi GPUs
name: Torch CUDA extension tests
needs: setup
if: contains(fromJson(needs.setup.outputs.matrix), 'deepspeed') || contains(fromJson(needs.setup.outputs.matrix), 'extended')
strategy:
fail-fast: false
matrix:
machines: [multi-gpu]
runs-on: [self-hosted, docker-gpu, '${{ matrix.machines }}']
machine_type: [multi-gpu]
runs-on: [self-hosted, docker-gpu, '${{ matrix.machine_type }}']
container:
image: nvcr.io/nvidia/pytorch:21.03-py3
options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
@@ -277,19 +277,19 @@ jobs:
- name: Run all non-slow selected tests on GPU
# TODO: Here we pass all tests in the 2 folders for simplicity. It's better to pass only the identified tests.
run: |
python -m pytest -n 1 --dist=loadfile -v --make-reports=${{ matrix.machines }}_tests_torch_cuda_extensions_gpu tests/deepspeed tests/extended
python -m pytest -n 1 --dist=loadfile -v --make-reports=${{ matrix.machine_type }}_tests_torch_cuda_extensions_gpu tests/deepspeed tests/extended
- name: Failure short reports
if: ${{ failure() }}
continue-on-error: true
run: cat reports/${{ matrix.machines }}_tests_torch_cuda_extensions_gpu/failures_short.txt
run: cat reports/${{ matrix.machine_type }}_tests_torch_cuda_extensions_gpu/failures_short.txt
- name: Test suite reports artifacts
if: ${{ always() }}
uses: actions/upload-artifact@v2
with:
name: ${{ matrix.machines }}_run_tests_torch_cuda_extensions_gpu_test_reports
path: reports/${{ matrix.machines }}_tests_torch_cuda_extensions_gpu
name: ${{ matrix.machine_type }}_run_tests_torch_cuda_extensions_gpu_test_reports
path: reports/${{ matrix.machine_type }}_tests_torch_cuda_extensions_gpu
send_results:
name: Send results to webhook