I wanted to experiment with Transfer Learning Mobilenet programming assignment on my own pc and download the class images. It is easy to do it one by one but it’s really time consuming.
Is there a way within the course platform to download the full dataset in one go, or at least all images within a specific folder?
Try to run this in a code cell (I use the first one with imports usually).
!tar chvfz notebook.tar.gz *
Then File → Open, and download the .tar.gz file.
You can insert cells in the notebook for things like that. Use “Insert → Cell Below”.
The suggestion of using tar is great, because that gives you all the files in one package. You may find that the resulting file is too large to download in one shot. Then you can use the linux “split” command. Here is a thread with more details on how to do that.
Note that if all you care about are the images, those are already packaged as a single h5 file that the logic of the notebook then reads and unpacks. You can just download the h5 file (also using tar as on that thread) and then use the python routine they provide for reading the images out of the file.
Also note that running things locally is a non-trivial exercise. We don’t have any comprehensive instructions for how to do that, but here is a thread from a fellow student with a lot of good information.
Thank you, I will give it a try
Thank you for suggestions. I will also review the thread from the other student