pytorch-widedeep icon indicating copy to clipboard operation
pytorch-widedeep copied to clipboard

The dataset is too large, the image processing is too slow, and the notebook crashes

Open xylovezxy opened this issue 9 months ago • 1 comments

HELLO,The datasets used by your columns are all small datasets, such as movelens100k. If I use the movelens1M dataset, the processing speed of image data will be slow and errors will eventually occur, possibly due to memory explosion. My notebook will crash. How to deal with this problem

xylovezxy avatar Feb 18 '25 10:02 xylovezxy

Hey @xylovezxy I will try to look into this asap.

Maybe @5uperpalo has some time?

jrzaurin avatar Feb 27 '25 11:02 jrzaurin