This repository is about my lessons learned during the Nanodegree. Feel free to look around!
If you are about to follow the course, it may help you to see a possible solution.
Do not copy and paste to pass the projects!
My favorite is the transfer learning part. I added a folder where tweaked code for GPU support can be found.
Moreover, I really enjoyed style transfer. An extra folder is added for this as well.
You can supply your images (style and target) and see what kind of art you can create 😃
- Create a virtual environment with all required libraries.
- Activate the environment.
- Start Jupiter notebook server.
You can do something like this:
conda env create -f environment.yml
conda activate deep-learning
jupyter notebook
You will find the environment.yml in this repo as well.
💻 Last tested 2022/06/05 on Windows 10 Home with Conda 4.12.0
Alternatively, you can also run a Jupyter notebook inside a container on a server (possibly with multiple GPUs). I assume you have docker installed. If you are new to Docker, you can start here.
- Upload the Dockerfile and requirement.txt to a server. Choose a folder to work in.
- Build Docker image.
- Run Docker image (starts jupyter server automatically).
- Open Jupyter in the browser.
You can do something like this when you are in the target folder on the server:
docker build -t deep-learning .
docker run -it --rm --gpus all --ipc=host -v $(pwd):/usr/src/workspace -p 9998:9998 --name deep-learning deep-learning
Go to your browser and connect to the Jupyter server with <ip-adress of your server>:9998.
Access is token-based. The token is displayed when you start the container. Look for the ?token= sequence and copy the token.
💻 Last tested 2022/06/05 on Ubuntu 18.04.6 LTS with Docker 20.10.14