Workshop

The examples cover a straightforward start, from shallow to intermediate, deep and CNN networks. It also shows how the trained model can be tested and evaluated.

Requirements

• Anaconda 3
• Docker Desktop

Build Docker Image

After cloning this repository, please execute the command below to build the Docker image.

docker build -t deeplearning-stack .

Run Docker Container

Once you have built the image, please execute the command below to run the container.

docker-compose up

• Remark: 'jovyan' is the default Docker user.

Jupyter Notebooks

After starting the Docker container, copy the Jupyter notebook URL and start working.

• Remark: if you face problems concerning lack of resources, please increase your Docker Engine memory. I tested the notebooks in a MacBook Pro with 16GB of RAM. I dedicated 5GB to my Docker Engine.

Running with Anaconda

The project also has a environment.yml file that can be used to create an Anaconda environment. One might be willing to use it instead of a docker container. To create, activate and run Jupyter Lab from the environment, check the commands below:

• conda env create -f environment.yml
  • This will create the dl-workshop environment.
• conda activate dl-workshop
  • This will activate the environment.
• jupyter-lab
  • This will start Jupyter Lab and open it in the browser.

TensorBoard

If you want to visualise the loss and accuracy metrics, just execute TensorBoard pointing to your logs directory:

tensorboard --logdir [path_to_project]/notebooks/logs

• Remark: the 'logs' directory is not part of the repository. It has to be created under the 'notebooks' directory. All the Jupyter notebook are already configured to use 'notebooks/logs' for the TensorBoard files.