Docker: TensorFlow with MIVisionX

Docker - TensorFlow + MIVisionX for Training

Steps to create docker for TensorFlow training with MIVisionX RALI:

1. Pull latest rocm/tensorflow docker

sudo docker pull rocm/tensorflow:rocm3.7-tf1.15-dev

2. Run all steps under Python Package Installation

cd models/research
# Compile protos.
protoc object_detection/protos/*.proto --python_out=.
# Install TensorFlow Object Detection API.
cp object_detection/packages/tf1/setup.py .
python -m pip install --use-feature=2020-resolver .

# Test the installation.
python object_detection/builders/model_builder_tf1_test.py

3. Clone MIVisionX. Run the setup script with '--neural_net off' option.

git clone https://github.com/GPUOpen-ProfessionalCompute-Libraries/MIVisionX
cd MIVisionX
python MIVisionX-setup.py --neural_net off

4. Build and install MIVisionX. Comment out lines for find_package - miopen and miopengemm - in MIVisionX/amd_openvx_extensions/CMakeLists.txt and MIVisionX/apps/CMakeLists.txt

cd MIVisionX
mkdir build
cd build
cmake ../
make -j
sudo make install

5. Setup rali_pybind

sudo ./run.sh

6. Install TensorFlow hub

pip3 isntall tensorflow_hub

7. Commit and save the new docker state

sudo docker commit <container ID> <new_image_name:tag_name>
sudo docker push <new_image_name:tag_name>