Files
transformers/.github/workflows/model_jobs.yml
陈赣 06f1fd69a6
Some checks failed
Self-hosted runner (nightly-past-ci-caller) / Get number (push) Has been cancelled
Self-hosted runner (nightly-past-ci-caller) / TensorFlow 2.11 (push) Has been cancelled
Self-hosted runner (nightly-past-ci-caller) / TensorFlow 2.10 (push) Has been cancelled
Self-hosted runner (nightly-past-ci-caller) / TensorFlow 2.9 (push) Has been cancelled
Self-hosted runner (nightly-past-ci-caller) / TensorFlow 2.8 (push) Has been cancelled
Self-hosted runner (nightly-past-ci-caller) / TensorFlow 2.7 (push) Has been cancelled
Self-hosted runner (nightly-past-ci-caller) / TensorFlow 2.6 (push) Has been cancelled
Self-hosted runner (nightly-past-ci-caller) / TensorFlow 2.5 (push) Has been cancelled
Self-hosted runner (benchmark) / Benchmark (aws-g5-4xlarge-cache) (push) Has been cancelled
Build documentation / build (push) Has been cancelled
Build documentation / build_other_lang (push) Has been cancelled
CodeQL Security Analysis / CodeQL Analysis (push) Has been cancelled
New model PR merged notification / Notify new model (push) Has been cancelled
PR CI / pr-ci (push) Has been cancelled
Slow tests on important models (on Push - A10) / Get all modified files (push) Has been cancelled
Secret Leaks / trufflehog (push) Has been cancelled
Update Transformers metadata / build_and_package (push) Has been cancelled
Slow tests on important models (on Push - A10) / Model CI (push) Has been cancelled
Check Tiny Models / Check tiny models (push) Has been cancelled
Self-hosted runner (Intel Gaudi3 scheduled CI caller) / Model CI (push) Has been cancelled
Self-hosted runner (Intel Gaudi3 scheduled CI caller) / Pipeline CI (push) Has been cancelled
Self-hosted runner (Intel Gaudi3 scheduled CI caller) / Example CI (push) Has been cancelled
Self-hosted runner (Intel Gaudi3 scheduled CI caller) / DeepSpeed CI (push) Has been cancelled
Self-hosted runner (Intel Gaudi3 scheduled CI caller) / Trainer/FSDP CI (push) Has been cancelled
Nvidia CI - Flash Attn / Setup (push) Has been cancelled
Nvidia CI - Flash Attn / Model CI (push) Has been cancelled
Nvidia CI / Setup (push) Has been cancelled
Nvidia CI / Model CI (push) Has been cancelled
Nvidia CI / Torch pipeline CI (push) Has been cancelled
Nvidia CI / Example CI (push) Has been cancelled
Nvidia CI / Trainer/FSDP CI (push) Has been cancelled
Nvidia CI / DeepSpeed CI (push) Has been cancelled
Nvidia CI / Quantization CI (push) Has been cancelled
Nvidia CI / Kernels CI (push) Has been cancelled
Doctests / Setup (push) Has been cancelled
Doctests / Call doctest jobs (push) Has been cancelled
Doctests / Send results to webhook (push) Has been cancelled
Extras Smoke Test / Get supported Python versions (push) Has been cancelled
Extras Smoke Test / Test extras on Python ${{ matrix.python-version }} (push) Has been cancelled
Extras Smoke Test / Check Slack token availability (push) Has been cancelled
Extras Smoke Test / Notify failures to Slack (push) Has been cancelled
Self-hosted runner (AMD scheduled CI caller) / Trigger Scheduled AMD CI (push) Has been cancelled
Stale Bot / Close Stale Issues (push) Has been cancelled
first commit
2026-06-05 16:53:03 +08:00

220 lines
8.3 KiB
YAML

name: model jobs
on:
workflow_call:
inputs:
folder_slices:
required: true
type: string
machine_type:
required: true
type: string
slice_id:
required: true
type: number
docker:
required: true
type: string
commit_sha:
required: false
type: string
report_name_prefix:
required: false
default: run_models_gpu
type: string
runner_type:
required: false
type: string
report_repo_id:
required: false
type: string
pytest_marker:
required: false
type: string
env:
HF_HOME: /mnt/cache
TRANSFORMERS_IS_CI: yes
OMP_NUM_THREADS: 8
MKL_NUM_THREADS: 8
RUN_SLOW: yes
# For gated repositories, we still need to agree to share information on the Hub repo. page in order to get access.
# This token is created under the bot `hf-transformers-bot`.
HF_TOKEN: ${{ secrets.HF_HUB_READ_TOKEN }}
TF_FORCE_GPU_ALLOW_GROWTH: true
CUDA_VISIBLE_DEVICES: 0,1
permissions:
contents: read
jobs:
run_models_gpu:
name: " "
strategy:
max-parallel: 8
fail-fast: false
matrix:
folders: ${{ fromJson(inputs.folder_slices)[inputs.slice_id] }}
runs-on:
group: '${{ inputs.machine_type }}'
container:
image: ${{ inputs.docker }}
options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
outputs:
machine_type: ${{ steps.set_machine_type.outputs.machine_type }}
steps:
- name: Echo input and matrix info
shell: bash
env:
folder_slices: ${{ inputs.folder_slices }}
matrix_folders: ${{ matrix.folders }}
slice_data: ${{ toJson(fromJson(inputs.folder_slices)[inputs.slice_id]) }}
run: |
echo "$folder_slices"
echo "$matrix_folders"
echo "$slice_data"
- name: Echo folder ${{ matrix.folders }}
shell: bash
# For folders like `models/bert`, set an env. var. (`matrix_folders`) to `models_bert`, which will be used to
# set the artifact folder names (because the character `/` is not allowed).
env:
matrix_folders_raw: ${{ matrix.folders }}
run: |
echo "$matrix_folders_raw"
matrix_folders="${matrix_folders_raw/'models/'/'models_'}"
echo "$matrix_folders"
echo "matrix_folders=$matrix_folders" >> $GITHUB_ENV
- name: Update clone
working-directory: /transformers
env:
commit_sha: ${{ inputs.commit_sha || github.sha }}
run: |
git fetch origin "$commit_sha" && git checkout "$commit_sha"
- name: Reinstall transformers in edit mode (remove the one installed during docker image build)
working-directory: /transformers
run: python3 -m pip uninstall -y transformers && python3 -m pip install -e .
- name: Update / Install some packages (for Past CI)
if: ${{ contains(inputs.docker, '-past-') }}
working-directory: /transformers
run: |
python3 -m pip install -U datasets
- name: Update / Install some packages (for Past CI)
if: ${{ contains(inputs.docker, '-past-') && contains(inputs.docker, '-pytorch-') }}
working-directory: /transformers
run: |
python3 -m pip install --no-cache-dir git+https://github.com/huggingface/accelerate@main#egg=accelerate
- name: NVIDIA-SMI
run: |
nvidia-smi
- name: Environment
working-directory: /transformers
run: |
python3 utils/print_env.py
- name: Show installed libraries and their versions
working-directory: /transformers
run: pip freeze
- name: Set `machine_type` for report and artifact names
id: set_machine_type
working-directory: /transformers
shell: bash
env:
input_machine_type: ${{ inputs.machine_type }}
run: |
echo "$input_machine_type"
if [ "$input_machine_type" = "aws-g5-4xlarge-cache" ]; then
machine_type=single-gpu
elif [ "$input_machine_type" = "aws-g5-12xlarge-cache" ]; then
machine_type=multi-gpu
else
machine_type="$input_machine_type"
fi
echo "$machine_type"
echo "machine_type=$machine_type" >> $GITHUB_ENV
echo "machine_type=$machine_type" >> $GITHUB_OUTPUT
- name: Create report directory if it doesn't exist
shell: bash
env:
report_name_prefix: ${{ inputs.report_name_prefix }}
run: |
mkdir -p "/transformers/reports/${machine_type}_${report_name_prefix}_${matrix_folders}_test_reports"
echo "dummy" > "/transformers/reports/${machine_type}_${report_name_prefix}_${matrix_folders}_test_reports/dummy.txt"
ls -la "/transformers/reports/${machine_type}_${report_name_prefix}_${matrix_folders}_test_reports"
- name: Run all tests on GPU
working-directory: /transformers
env:
report_name_prefix: ${{ inputs.report_name_prefix }}
pytest_marker: ${{ inputs.pytest_marker }}
model: ${{ matrix.folders }}
run: |
# Map short names to actual test paths for trainer/distributed tests
test_path="tests/${model}"
if [ "$model" = "fsdp" ]; then
test_path="tests/trainer/distributed/test_trainer_distributed_fsdp.py"
elif [ "$model" = "ddp" ]; then
test_path="tests/trainer/distributed/test_trainer_distributed_ddp.py"
elif [ "$model" = "trainer" ] && [ "$report_name_prefix" = "run_trainer_and_fsdp_gpu" ]; then
test_path="tests/trainer --ignore=tests/trainer/distributed"
fi
script -q -c "PATCH_TESTING_METHODS_TO_COLLECT_OUTPUTS=yes _PATCHED_TESTING_METHODS_OUTPUT_DIR=/transformers/reports/${machine_type}_${report_name_prefix}_${matrix_folders}_test_reports python3 -m pytest -rsfE -v -m '${pytest_marker}' --make-reports=${machine_type}_${report_name_prefix}_${matrix_folders}_test_reports ${test_path}" test_outputs.txt
ls -la
# Extract the exit code from the output file
EXIT_CODE=$(tail -1 test_outputs.txt | grep -o 'COMMAND_EXIT_CODE="[0-9]*"' | cut -d'"' -f2)
exit ${EXIT_CODE:-1}
- name: Failure short reports
if: ${{ failure() }}
# This step is only to show information on Github Actions log.
# Always mark this step as successful, even if the report directory or the file `failures_short.txt` in it doesn't exist
continue-on-error: true
env:
report_name_prefix: ${{ inputs.report_name_prefix }}
run: cat "/transformers/reports/${machine_type}_${report_name_prefix}_${matrix_folders}_test_reports/failures_short.txt"
- name: Captured information
if: ${{ failure() }}
continue-on-error: true
env:
report_name_prefix: ${{ inputs.report_name_prefix }}
run: |
cat "/transformers/reports/${machine_type}_${report_name_prefix}_${matrix_folders}_test_reports/captured_info.txt"
- name: Copy test_outputs.txt
if: ${{ always() }}
continue-on-error: true
env:
report_name_prefix: ${{ inputs.report_name_prefix }}
run: |
cp /transformers/test_outputs.txt "/transformers/reports/${machine_type}_${report_name_prefix}_${matrix_folders}_test_reports"
- name: "Test suite reports artifacts: ${{ env.machine_type }}_${{ inputs.report_name_prefix }}_${{ env.matrix_folders }}_test_reports"
if: ${{ always() }}
uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
with:
name: ${{ env.machine_type }}_${{ inputs.report_name_prefix }}_${{ env.matrix_folders }}_test_reports
path: /transformers/reports/${{ env.machine_type }}_${{ inputs.report_name_prefix }}_${{ env.matrix_folders }}_test_reports
collated_reports:
name: Collated Reports
if: ${{ always() && inputs.runner_type != '' }}
needs: run_models_gpu
uses: huggingface/transformers/.github/workflows/collated-reports.yml@6abd9725ee7d809dc974991f8ff6c958afb63a3a # main
with:
job: run_models_gpu
report_repo_id: ${{ inputs.report_repo_id }}
gpu_name: ${{ inputs.runner_type }}
machine_type: ${{ needs.run_models_gpu.outputs.machine_type }}
secrets: inherit