zerotosingularity/fastai repository overview

⁠Important

This container has moved to https://hub.docker.com/repository/docker/seemeai/fastai2⁠.

The existing tags will remain for a while, but we recommend switching to using https://hub.docker.com/repository/docker/seemeai/fastai2⁠.

There you will find:

• newer versions of the framework
• support for different PyTorch/torchvision versions

⁠Old info for reference

⁠DISCLAIMER

This software comes as-is and with no warranty whatsoever. You are free to use this container at your own risk and for your own benefit.

⁠How to use this container

1. Install Docker: (docker-ce or docker-ee) https://docs.docker.com/install/⁠

2. Install Nvidia-docker: https://github.com/NVIDIA/nvidia-docker⁠

3. Checkout code and run container

git clone https://github.com/fastai/course-v3.git
cd course-v3
docker run --runtime=nvidia -it -p 8888:8888 -v $HOME/.fastai:/root/.fastai -v $(pwd):/code/fastai --ipc=host zerotosingularity/fastai:latest

1. Kaggle (Optional) If you want to use the Kaggle command line tools:

• place the kaggle.json file in the ~/.kaggle/ folder
• add another volume when starting the container: -v ~/.kaggle:/YOUR_USER/.kaggle
  • YOUR_USER should be changed to your actual username, because the target volume needs to be an absolute path
  • full example: docker run --runtime=nvidia -it -p 8888:8888 -v $HOME/.fastai:/root/.fastai -v $(pwd):/code/fastai -v ~/.kaggle:/YOUR_USER/.kaggle --ipc=host zerotosingularity/fastai:latest

It uses three volumes:

• /root/.fastai: store the data for later use, so you don't have to redownload every time
• /root/.torchi: store the data for later use, so you don't have to redownload every time
• /code/fastai: maps to the current (course-v3) repository, which lets you save changes over time, and simply pull updates
• (Optional Kaggle volume as described in 4. Kaggle (Optional))