I have .csv file in which i have around 50000 image urls how to load these image dataset.
Please help.
more information required.
What have you got? (e.g. show 20 lines of .csv)
What is the end result you want? (where will they go, directory structure, labels etc
As Jeremy says in a lecture somewhere, its probably less work to make the data suit an existing data loader then it is to write a new one
i want to load this data and task is to classify these into given Sub category
lookup urllib.request.urlretrieve (google, stackoverflow etc) for the downloading part. You’ll need to decide what to do with the files in your code
here’s a fastai thread thats a little related
1 Like
I was able to download the image files for a list of urls using $ wget -i list_filename
If I were you, I would select that whole column of URLs and paste it into a different text file. Then I would cd to that directory in the terminal and run that wget, where the name of that text file goes in place of list_filename.
1 Like
but then how would i classify them in given categories
Could you sort by category first and then copy the URLs for each category into the wget command separately?
Yes ,i have done that.