Platform: Colab ✅

how do i back up my project data and models?

go to file -> save a copy in drive

i don’t understand why it keeps getting interrupted. I’ve tried running it multiple times but it still the same.

1 Like

It is normal but you can change this behavior by putting lr_find(stop_div=False).
All information available in the docs


This will only save a copy of the notebook not the files or the environment.

The best way right now is to tarball all data and download it or transfer it to your google drive.


How to do this @gamo ?
How to upload the bear files ? I ran all the commands for the 3 categories separately but where am I going to find the upload button?

1 Like

Simple way to upload is to use the gui for small files.



The best way to save lots of data is to mount your google drive.

from google.colab import *
drive.mount('/content/gdrive', force_remount=True)

You will have to authenticate by following the link, choosing google account, authenticate, copy code and pasting it in the notebook. After that the google drive will be mounted in file system.


Thank u it worked

do we have to do this for every notebook?

Yes, and everytime you reset the runtime.

some more help pls, how to delete the files uploaded in working directory? I don’t see any delete option

1 Like

click on file, then move it to trash

r u sure, isn’t it for notebook? I’m trying to delete the data downloaded?

i thought it is for the notebook :grin:. I don’t have any idea of how to delete the files.

You will have to use !rm to delete files with -r switch to recursively delete.

To delete all data !rm -r /content/data/*


I think @jeremy will explain us how learning rate finder works in the forthcoming lessons. For now, know that learning rate finder gets interrupted when the a particular learning rate gives too high a loss for it to be useful, and it is the time when the learning rate finder must stop. Hence interruption for lr_find() is nothing unnatural; you can even see that there is a message called LR Finder complete, type {learner_name}.recorder.plot() to see the graph. Hence nothing is wrong.

1 Like

Is doc() working for everyone in Colab?? Or is it just me… :frowning:

1 Like

me too!

ok, thanks