transformers/.github/workflows/self-push.yml

name: Self-hosted runner (push)

on:
  push:
    branches:
      - master
      - ci_*
    paths:
      - "src/**"
      - "tests/**"
      - ".github/**"
      - "templates/**"
  # pull_request:
  repository_dispatch:


jobs:
  run_tests_torch_gpu:
    runs-on: [self-hosted, gpu, single-gpu]
    steps:
      - uses: actions/checkout@v2
      - name: Python version
        run: |
          which python
          python --version
          pip --version          

      - name: Current dir
        run: pwd

      - run: nvidia-smi

      - name: Kill any run-away pytest processes
        run: pkill -f tests; pkill -f examples

      - name: Loading cache.
        uses: actions/cache@v2
        id: cache
        with:
          path: .env
          key: v1.1-tests_torch_gpu-${{ hashFiles('setup.py') }}

      - name: Create new python env (on self-hosted runners we have to handle isolation ourselves)
        run: |
          python -m venv .env
          source .env/bin/activate
          which python
          python --version
          pip --version          

      - name: Install dependencies
        run: |
          source .env/bin/activate
          pip install --upgrade pip
          pip install .[torch,sklearn,testing,onnxruntime,sentencepiece]
          pip install git+https://github.com/huggingface/datasets
          pip install pandas torch-scatter -f https://pytorch-geometric.com/whl/torch-1.7.0+cu102.html          

      - name: Are GPUs recognized by our DL frameworks
        run: |
          source .env/bin/activate
          python -c "import torch; print('Cuda available:', torch.cuda.is_available())"
          python -c "import torch; print('Number of GPUs available:', torch.cuda.device_count())"          

#      - name: Create model files
#        run: |
#          source .env/bin/activate
#          transformers-cli add-new-model --testing --testing_file=templates/adding_a_new_model/tests/encoder-bert-tokenizer.json --path=templates/adding_a_new_model
#          transformers-cli add-new-model --testing --testing_file=templates/adding_a_new_model/tests/pt-encoder-bert-tokenizer.json --path=templates/adding_a_new_model
#          transformers-cli add-new-model --testing --testing_file=templates/adding_a_new_model/tests/standalone.json --path=templates/adding_a_new_model
#          transformers-cli add-new-model --testing --testing_file=templates/adding_a_new_model/tests/tf-encoder-bert-tokenizer.json --path=templates/adding_a_new_model

      - name: Run all non-slow tests on GPU
        env:
          OMP_NUM_THREADS: 1
          CUDA_VISIBLE_DEVICES: 0
        run: |
          source .env/bin/activate
          python -m pytest -n 2 --dist=loadfile -s --make-reports=tests_torch_gpu tests          

      - name: Failure short reports
        if: ${{ always() }}
        run: cat reports/tests_torch_gpu_failures_short.txt

      - name: Test suite reports artifacts
        if: ${{ always() }}
        uses: actions/upload-artifact@v2
        with:
          name: run_all_tests_torch_gpu_test_reports
          path: reports


  run_tests_tf_gpu:
    runs-on: [self-hosted, gpu, single-gpu]
    steps:
      - uses: actions/checkout@v2
      - name: Python version
        run: |
          which python
          python --version
          pip --version          

      - name: Current dir
        run: pwd

      - run: nvidia-smi

      - name: Kill any run-away pytest processes
        run: pkill -f tests; pkill -f examples

      - name: Loading cache.
        uses: actions/cache@v2
        id: cache
        with:
          path: .env
          key: v1.1-tests_tf_gpu-${{ hashFiles('setup.py') }}

      - name: Create new python env (on self-hosted runners we have to handle isolation ourselves)
        run: |
          python -m venv .env
          source .env/bin/activate
          which python
          python --version
          pip --version          

      - name: Install dependencies
        run: |
          source .env/bin/activate
          pip install --upgrade pip
          pip install .[tf,sklearn,testing,onnxruntime,sentencepiece]
          pip install git+https://github.com/huggingface/datasets          

      - name: Are GPUs recognized by our DL frameworks
        run: |
          source .env/bin/activate
          TF_CPP_MIN_LOG_LEVEL=3 python -c "import tensorflow as tf; print('TF GPUs available:', bool(tf.config.list_physical_devices('GPU')))"
          TF_CPP_MIN_LOG_LEVEL=3 python -c "import tensorflow as tf; print('Number of TF GPUs available:', len(tf.config.list_physical_devices('GPU')))"          

      - name: Create model files
        run: |
          source .env/bin/activate          
#          transformers-cli add-new-model --testing --testing_file=templates/adding_a_new_model/tests/encoder-bert-tokenizer.json --path=templates/adding_a_new_model
#          transformers-cli add-new-model --testing --testing_file=templates/adding_a_new_model/tests/pt-encoder-bert-tokenizer.json --path=templates/adding_a_new_model
#          transformers-cli add-new-model --testing --testing_file=templates/adding_a_new_model/tests/standalone.json --path=templates/adding_a_new_model
#          transformers-cli add-new-model --testing --testing_file=templates/adding_a_new_model/tests/tf-encoder-bert-tokenizer.json --path=templates/adding_a_new_model

      - name: Run all non-slow tests on GPU
        env:
          OMP_NUM_THREADS: 1
          CUDA_VISIBLE_DEVICES: 0
        run: |
          source .env/bin/activate
          python -m pytest -n 2 --dist=loadfile -s --make-reports=tests_tf_gpu tests          

      - name: Failure short reports
        if: ${{ always() }}
        run: cat reports/tests_tf_gpu_failures_short.txt

      - name: Test suite reports artifacts
        if: ${{ always() }}
        uses: actions/upload-artifact@v2
        with:
          name: run_all_tests_tf_gpu_test_reports
          path: reports

  run_tests_torch_multi_gpu:
    runs-on: [self-hosted, gpu, multi-gpu]
    steps:
      - uses: actions/checkout@v2
      - name: Python version
        run: |
          which python
          python --version
          pip --version          

      - name: Current dir
        run: pwd

      - run: nvidia-smi

      - name: Kill any run-away pytest processes
        run: pkill -f tests; pkill -f examples

      - name: Loading cache.
        uses: actions/cache@v2
        id: cache
        with:
          path: .env
          key: v1.1-tests_torch_multi_gpu-${{ hashFiles('setup.py') }}

      - name: Create new python env (on self-hosted runners we have to handle isolation ourselves)
        run: |
          python -m venv .env
          source .env/bin/activate
          which python
          python --version
          pip --version          
      - name: Install dependencies
        run: |
          source .env/bin/activate
          pip install --upgrade pip
          pip install .[torch,sklearn,testing,onnxruntime,sentencepiece]
          pip install git+https://github.com/huggingface/datasets
          pip install pandas torch-scatter -f https://pytorch-geometric.com/whl/torch-1.7.0+cu102.html          

      - name: Are GPUs recognized by our DL frameworks
        run: |
          source .env/bin/activate
          python -c "import torch; print('Cuda available:', torch.cuda.is_available())"
          python -c "import torch; print('Number of GPUs available:', torch.cuda.device_count())"          

      - name: Run all non-slow tests on GPU
        env:
          OMP_NUM_THREADS: 1
        run: |
          source .env/bin/activate
          python -m pytest -n 2 --dist=loadfile -s --make-reports=tests_torch_multi_gpu tests          

      - name: Failure short reports
        if: ${{ always() }}
        run: cat reports/tests_torch_multi_gpu_failures_short.txt

      - name: Test suite reports artifacts
        if: ${{ always() }}
        uses: actions/upload-artifact@v2
        with:
          name: run_all_tests_torch_multi_gpu_test_reports
          path: reports

  run_tests_tf_multi_gpu:
    runs-on: [self-hosted, gpu, multi-gpu]
    steps:
      - uses: actions/checkout@v2
      - name: Python version
        run: |
          which python
          python --version
          pip --version          

      - name: Current dir
        run: pwd

      - run: nvidia-smi

      - name: Kill any run-away pytest processes
        run: pkill -f tests; pkill -f examples

      - name: Loading cache.
        uses: actions/cache@v2
        id: cache
        with:
          path: .env
          key: v1.1-tests_tf_multi_gpu-${{ hashFiles('setup.py') }}

      - name: Create new python env (on self-hosted runners we have to handle isolation ourselves)
        run: |
          python -m venv .env
          source .env/bin/activate
          which python
          python --version
          pip --version          
      - name: Install dependencies
        run: |
          source .env/bin/activate
          pip install --upgrade pip
          pip install .[tf,sklearn,testing,onnxruntime,sentencepiece]
          pip install git+https://github.com/huggingface/datasets          

      - name: Are GPUs recognized by our DL frameworks
        run: |
          source .env/bin/activate
          TF_CPP_MIN_LOG_LEVEL=3 python -c "import tensorflow as tf; print('TF GPUs available:', bool(tf.config.list_physical_devices('GPU')))"
          TF_CPP_MIN_LOG_LEVEL=3 python -c "import tensorflow as tf; print('Number of TF GPUs available:', len(tf.config.list_physical_devices('GPU')))"          

      - name: Run all non-slow tests on GPU
        env:
          OMP_NUM_THREADS: 1
        run: |
          source .env/bin/activate
          python -m pytest -n 2 --dist=loadfile -s --make-reports=tests_tf_multi_gpu tests          

      - name: Failure short reports
        if: ${{ always() }}
        run: cat reports/tests_tf_multi_gpu_failures_short.txt

      - name: Test suite reports artifacts
        if: ${{ always() }}
        uses: actions/upload-artifact@v2
        with:
          name: run_all_tests_tf_multi_gpu_test_reports
          path: reports