Github Actions workflow CI for horovod on CPU

.github/workflows/hvd-tests.yml

@@ -0,0 +1,99 @@
+name: Run Horovod tests
+on: [push, pull_request]
+
+jobs:
+  horovod-tests:
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        python-version: [3.7]
+        pytorch-channel: [pytorch]
+
+
+    steps:
+    - uses: actions/checkout@v2
+    - name: Setup Miniconda
+      uses: goanpeca/setup-miniconda@v1
+      with:
+        miniconda-version: "latest"
+        python-version: ${{ matrix.python-version }}
+
+    - name: Install dependencies
+      shell: bash -l {0}
+      run: |      
+        #install other dependencies
+        conda install pytorch torchvision cpuonly -c ${{ matrix.pytorch-channel }}
+        pip install -r requirements-dev.txt
+        # Fixes #1153
+        pip install --upgrade scipy==1.4.1
+        pip install horovod
+        python setup.py install
+
+    - name: Run Tests
+      shell: bash -l {0}
+      run: |
+        SKIP_DISTRIB_TESTS=${{ matrix.skip-distrib-tests }} CI_PYTHON_VERSION="${{ matrix.python-version }}" sh tests/run_cpu_tests.sh
+
+    - name: Run MNIST Examples
+      shell: bash -l {0}
+      run: |
+        # MNIST
+        # 1) mnist.py
+        python examples/mnist/mnist.py --epochs=1
+
+    - name: Run MNIST with loggers Examples
+      shell: bash -l {0}
+      if: ${{ matrix.os == 'ubuntu-latest' }}
+      run: |
+        # 2) mnist_with_visdom.py
+        python -c "from visdom.server import download_scripts; download_scripts()" # download scripts : https://github.com/facebookresearch/visdom/blob/master/py/server.py#L929
+        python -m visdom.server &
+        sleep 10
+        python examples/mnist/mnist_with_visdom.py --epochs=1
+        kill %1
+        # 3.1) mnist_with_tensorboard.py with tbX
+        python examples/mnist/mnist_with_tensorboard.py --epochs=1
+        # 3.2) mnist_with_tensorboard.py with native torch tb
+        pip uninstall -y tensorboardX
+        python examples/mnist/mnist_with_tensorboard.py --epochs=1
+
+    - name: Run MNIST Example With Crash
+      shell: bash -l {0}
+      if: ${{ matrix.os == 'ubuntu-latest' }}
+      continue-on-error: true
+      run: |
+        # 4) mnist_save_resume_engine.py
+        python examples/mnist/mnist_save_resume_engine.py --epochs=2 --crash_iteration 1100
+
+    - name: Resume MNIST from previous crash
+      shell: bash -l {0}
+      if: ${{ matrix.os == 'ubuntu-latest' }}
+      run: |
+        python examples/mnist/mnist_save_resume_engine.py --epochs=2 --resume_from=/tmp/mnist_save_resume/checkpoint_1.pt
+
+    - name: Run GAN example
+      shell: bash -l {0}
+      if: ${{ matrix.os == 'ubuntu-latest' }}
+      run: |
+        # DCGAN
+        python examples/gan/dcgan.py --dataset fake --dataroot /tmp/fakedata --output-dir /tmp/outputs-dcgan --batch-size 2 --epochs 2  --workers 0
+
+    - name: Run RL Examples
+      shell: bash -l {0}
+      if: ${{ matrix.os == 'ubuntu-latest' }}
+      run: |
+        # RL
+        # 1) Actor-Critic
+        python examples/reinforcement_learning/actor_critic.py --max-episodes=2
+        # 2) Reinforce
+        python examples/reinforcement_learning/reinforce.py --max-episodes=2
+
+    - name: Run Neural Style Example
+      shell: bash -l {0}
+      if: ${{ matrix.os == 'ubuntu-latest' }}
+      run: |
+        #fast-neural-style
+        #train
+        mkdir -p ~/.cache/torch/checkpoints/ && wget "https://download.pytorch.org/models/vgg16-397923af.pth" -O ~/.cache/torch/checkpoints/vgg16-397923af.pth
+        python examples/fast_neural_style/neural_style.py train --epochs 1 --cuda 0 --dataset test --dataroot . --image_size 32 --style_image examples/fast_neural_style/images/style_images/mosaic.jpg --style_size 32
+