I am using ResNet-18 for classification purpose. I have used dataloader to load the data. But how get the label to class name mapping? Does it load in alphabetical order?
For future reference:
I have used this from https://github.com/pytorch/vision/blob/master/torchvision/datasets/folder.py to find the mapping-
def find_classes(dir):
classes = os.listdir(dir)
classes.sort()
class_to_idx = {classes[i]: i for i in range(len(classes))}
return classes, class_to_idx
@avijit_dasgupta hi, I have the same problem. Can you provide a class_idx to class name mapping? That would be super convenient.
Something like this:
dataset = datasets.ImageFolder(data_folder, transform=data_transforms)
print(dataset.class_to_idx)
10 Likes
For idx to classes mappping, just get list of keys from dict returned from dataset.class_to_idx
1 Like