Hello guys!, i’ve just started the course, and started using Colab for the first time. I’m trying to experiment with the data sets available at fastai but for some reason they don’t work what am i doing wrong. Thanks in advance.
You should grab the fastai datasets via untar_data and using the URLs.dataset option (press tab after the URLs.
For the actual untarring, you should use either the Linux tar command (google that, it’s a bash command 
).
1 Like
Thanks a lot it works now!. But how do i point it to a local colab file that i made ? Thanks!
Check out what the path is when you do path = untar_data(URLs.X)
That path is where it’s stored
AttributeError Traceback (most recent call last)
in ()
7
8
----> 9 path = untar_data(URLs.x)
10 print(path)
AttributeError: type object ‘URLs’ has no attribute ‘x’
I meant the X as example. There’s URLs.ADULT_SAMPLE for the salary dataset, PETS for the dogs and cats dataset, etc. if you press Tab after typing URLs. It’ll show the entire list
1 Like
I figured that one out, but instead of using fastai datasets , i’m trying to use it on 50 pictures of motorcycles that i’ve put together to see what happens.
You should follow lesson 2 then, as Jeremy discusses this 
(in practical deep learning for coders)
Oh my bad im so sorry ! Just finished part 1, he says at the end use your own pictures just trying to learn everything hehe. Thanks again !
1 Like