You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
HI @theanswer003 ,
Thank you for the nice work. When I run the train.py of CycleGAN, I encountered the following issue with importing the dataset using the dataloader. Despite changing my dataset, I am consistently getting the following issue:
Traceback (most recent call last): File "train.py", line 78, in <module> dataloader = DataLoader(ImageDataset(opt.dataroot, transform=transform, unaligned=True), File "/opt/conda/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 347, in __init__ sampler = RandomSampler(dataset, generator=generator) # type: ignore[arg-type] File "/opt/conda/lib/python3.8/site-packages/torch/utils/data/sampler.py", line 107, in __init__ raise ValueError("num_samples should be a positive integer " ValueError: num_samples should be a positive integer value, but got num_samples=0
best regards,
A
The text was updated successfully, but these errors were encountered:
HI @theanswer003 ,
Thank you for the nice work. When I run the train.py of CycleGAN, I encountered the following issue with importing the dataset using the dataloader. Despite changing my dataset, I am consistently getting the following issue:
Traceback (most recent call last): File "train.py", line 78, in <module> dataloader = DataLoader(ImageDataset(opt.dataroot, transform=transform, unaligned=True), File "/opt/conda/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 347, in __init__ sampler = RandomSampler(dataset, generator=generator) # type: ignore[arg-type] File "/opt/conda/lib/python3.8/site-packages/torch/utils/data/sampler.py", line 107, in __init__ raise ValueError("num_samples should be a positive integer " ValueError: num_samples should be a positive integer value, but got num_samples=0
best regards,
A
The text was updated successfully, but these errors were encountered: