Using Chainer on a Mac is not supported, and I decided to use Docker because of the hassle of porting it when I want to run it on a server after Iāve created something.
$ brew install docker
$ brew cask install docker
$ docker run -p 8888:8888 --name ml -it asashiho/ml-jupyter-python3
- Docker is useful when studying machine learning - Asaās Alone
- https://hub.docker.com/r/asashiho/ml-jupyter-python3/~/dockerfile/
- When executed, Jupyter Notebook starts up.
- After Ctrl-C, the container does not disappear, but is stopped.
$ docker start /ml
$ docker exec /ml pwd
- /notebooks
If you want to access data files, results files, etc. from the outside, Bind mounts instead of Volumes seems to be the way to go.
$ docker run -p 8888:8888 --name ml -it --mount type=bind,source=/Users/nishio,target=/notebooks/home asashiho/ml-jupyter-python3
memo
- Chainer is not included.
- pip is old
- You should consider upgrading via the āpip install āupgrade pipā command.
-
pip install āupgrade pip
- Iāll be about 30% slower.
This page is auto-translated from [/nishio/Docker for Mac](https://scrapbox.io/nishio/Docker for Mac) using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. Iām very happy to spread my thought to non-Japanese readers.