0

Apologies if this has been asked before, I've been reading about this issue for a while and all solutions seem similar to the same 3 available options.

For example in this thread Load local data files to Colaboratory

It is explained how to manually upload a file. That could work. But what if we had to share a file with 100 users? I believe that with that type of solutions, they would all have to copy the collaboratory project to their local machine, as usual, and everyone would have to upload the same file, and then proceed to use the code, once they have the required CSV.

Am I missing something? Is there any way to share both the project and a number of files (even if that implies a longer load time since the file has to be downloaded for ever person, but I would like to know if there is an automatic way to do this).

Particularly interested in solutions that do not involve auth tokens since the people that would get the shared project are not technologically shavy enough for the task.

monkey intern
  • 705
  • 3
  • 14
  • 34

1 Answers1

1

If the data is not secret, git may be the best solution. You upload those csv files to github. Then use git clone in your Colab notebook.

korakot
  • 37,818
  • 16
  • 123
  • 144
  • I read somewhere git only lets you upload files up to a certain file size, something like 25MB might be? Any way to get 100-200-300MB files this way? – monkey intern Apr 29 '19 at 07:19
  • 1
    I saw somewhere they split the a big file into 4 files. Then merge them back later. But I forget where or how. – korakot Apr 29 '19 at 07:22