I am currently working with Flickr8k dataset. I downloaded the dataset from here. Then , I extracted the zip such that all Images are Flickr8k/Images and caption is in Flickr8k/captions.txt. Then I used the following code
from torch.utils.data import Dataset, DataLoader image_folder = './Flickr8k/Images' caption_path = './Flickr8k/captions.txt' data_transform = transforms.Compose([ transforms.Resize(128), transforms.ToTensor() ]) trainset = datasets.Flickr8k(root=image_folder, ann_file=caption_path, transform=data_transform)
The trainset does not load anything.
Dataset Flickr8k Number of datapoints: 0 Root location: ./Flickr8k/Images StandardTransform Transform: Compose( Resize(size=128, interpolation=bilinear, max_size=None, antialias=warn) ToTensor() )
Please tell me where I am going wrong. The Documentation does not specify anything.