ci/cd: drop horovod

* install sudo
* req
* hot: torchtext

Author: Borda

.azure/ipynb-publish.yml

@@ -23,7 +23,7 @@ jobs:
       # base ML image: mcr.microsoft.com/azureml/openmpi3.1.2-cuda10.2-cudnn8-ubuntu18.04
       image: "pytorchlightning/pytorch_lightning:base-cuda-py3.9-torch1.10"
       # image: "pytorch/pytorch:1.7.1-cuda11.0-cudnn8-runtime"
-      options: "-it --rm --runtime=nvidia -e NVIDIA_VISIBLE_DEVICES=all --shm-size=32g"
+      options: "--runtime=nvidia -e NVIDIA_VISIBLE_DEVICES=all --shm-size=32g --name cd-container -v /usr/bin/docker:/tmp/docker:ro"
 
     variables:
       ACCELERATOR: CPU,GPU
@@ -40,6 +40,11 @@ jobs:
         python --version
       displayName: 'Image info & NVIDIA'
 
+    - script: |
+        /tmp/docker exec -t -u 0 cd-container \
+        sh -c "apt-get update && DEBIAN_FRONTEND=noninteractive apt-get -o Dpkg::Options::="--force-confold" -y install sudo"
+      displayName: 'Install Sudo in container (thanks Microsoft!)'
+
     - bash: |
         git fetch --all
         echo $(PUB_BRANCH)
@@ -55,9 +60,12 @@ jobs:
       displayName: 'Set Git user'
 
     - bash: |
+        set -e
         pip --version
         pip install --requirement requirements.txt
         pip install --requirement requirements/data.txt
+        sudo pip uninstall -y horovod
+        sudo pip uninstall -y torchtext  # FIXME: this is hotfix for https://github.com/PyTorchLightning/pytorch-lightning/pull/12676
         pip list
       displayName: 'Install dependencies'
 

.azure/ipynb-tests.yml

@@ -19,7 +19,7 @@ jobs:
       # base ML image: mcr.microsoft.com/azureml/openmpi3.1.2-cuda10.2-cudnn8-ubuntu18.04
       image: "pytorchlightning/pytorch_lightning:base-cuda-py3.9-torch1.10"
       # image: "pytorch/pytorch:1.7.1-cuda11.0-cudnn8-runtime"
-      options: "-it --rm --runtime=nvidia -e NVIDIA_VISIBLE_DEVICES=all --shm-size=32g"
+      options: "--runtime=nvidia -e NVIDIA_VISIBLE_DEVICES=all --shm-size=32g --name ci-container -v /usr/bin/docker:/tmp/docker:ro"
 
     variables:
       ACCELERATOR: CPU,GPU
@@ -34,10 +34,18 @@ jobs:
         python --version
       displayName: 'Image info & NVIDIA'
 
+    - script: |
+        /tmp/docker exec -t -u 0 ci-container \
+        sh -c "apt-get update && DEBIAN_FRONTEND=noninteractive apt-get -o Dpkg::Options::="--force-confold" -y install sudo"
+      displayName: 'Install Sudo in container (thanks Microsoft!)'
+
     - bash: |
+        set -e
         pip --version
         pip install --requirement requirements.txt
         pip install --requirement requirements/data.txt
+        sudo pip uninstall -y horovod
+        sudo pip uninstall -y torchtext  # FIXME: this is hotfix for https://github.com/PyTorchLightning/pytorch-lightning/pull/12676
         pip list
       displayName: 'Install dependencies'
 

lightning_examples/reinforce-learning-DQN/.meta.yml

@@ -13,9 +13,6 @@ description: |
   2. Handle unsupervised learning by using an IterableDataset where the dataset itself is constantly updated during training
   3. Each training step carries has the agent taking an action in the environment and storing the experience in the IterableDataset
 requirements:
-  - torchvision<=0.10
-  - torchaudio<=0.10
-  - torchtext<=0.10
   - gym
   - pygame
 accelerator: