Hi,
I wanted to run the Jupyter notebooks locally on my laptop and was wondering if a Docker image is available that caters to the entire course or are there separate images available for each course.
If there are instructions to build a Docker image, that would be fine too.
Thanks,
Viji Sarathy
1 Like
Hi!
As far as I know, we don’t have docker images for the Jupyter notebooks that are available to be passed out to students.
I can bring this up as something to be looked into in the future.
Thanks for the suggestion!
Thanks.
Is it possible for someone to provide the list of all dependencies to be installed on a Linux server before I can run the notebooks that are used here in this class.
I could use those to create a Docker image myself and would be happy to share if anyone else is interested.
Bumping this thread up as theDocker images I could find (e.g. jupyter/tensorflow-notebook) don’t seem to work (import tensorflow_data_validation as tfdv triggers a missing module error for instance).
Maybe the image is not up to date. If you wish, you can try installing the module manually inside the image, but there is a possibility of encountering version conflicts.
The alternative strategy is to download all the files associated with the assignments to your computer and then install the relevant software packages. Your “missing module” error is simply that the notebooks themselves are not sufficient: each one is accompanied by a tree of other files. It turns out to be easy to download all the files for a given assignment, but getting all the required software installed and avoiding “versionitis” problems is not quite so easy. There are no official instructions for that, but here is a thread that will get you started down that path. Some problem solving skills will be required on your part, but if you’re going to be playing this game on a long term basis, the sooner you start developing (or further strengthening as the case may be) those skills the happier you will be.