Open Amirhossein-kargar-khabbazi opened 4 months ago
Hi, it looks like deeplake changed something in their new version. Could you please try running the original code with
pip install deeplake==3.8.21
We will look into this issue in detail soon.
Thank you for your reply, I tried this version but could not install it, I tried another version but still the problem is still there. I would be grateful if any of the members came over of this problem, let me know. Thank you
I used !pip install deeplake==3.8.21, but still I have this problem:
the deadline is close and we still couldn't pass the per-process of dataset due to dataset issue. we would be grateful if you fix it as soon as possible. Thanks
Hello, I was trying to run the exact code without any modification, but I saw the accuracy is 0.0, I understand that the data has a problem so I did some logs below and the labels were not 0-199. As you see in the screenshot below, there are data and labels.
Also during training, I did some logs to see the predicted and true labels, and the reason is that label distributions are not correct.
Also the results:
Thank you