admin管理员组

文章数量:1389196

Can somebody please explain me the file structure of downloaded MNIST datase from Kaggle, via kagglehub as zip(23MB) file, when unzip I get these:

  1. folder: t10k-images-idx3-ubyte --> (inside the folder) --> t10k-images-idx3-ubyte
  2. folder: t10k-images-idx1-ubyte --> (inside the folder) --> t10k-images-idx3-ubyte
  3. folder: train-images-idx3-ubyte --> (inside the folder) --> train-images-idx3-ubyte
  4. folder: train-images-idx1-ubyte --> (inside the folder) --> train-images-idx1-ubyte
  5. file: t10k-images.idx3-ubyte
  6. file: t10k-labels.idx1-ubyte
  7. file: train-images.idx3-ubyte
  8. file: train-labels.idx1-ubyte

no gz extensions... ...I would like to know if it is right the following procedure for reading the sets locally using pytorch:

import torch
from torchvision import datasets, transforms

transform = transforms.Compose([transforms.ToTensor()])

mnist_pytorch = datasets.MNIST(root='../', train=True, download=False, transform=transform)

thanks a lot for your time

本文标签： MNIST dataset file structureStack Overflow

版权声明：本文标题：MNIST dataset file structure - Stack Overflow 内容由网友自发贡献，该文观点仅代表作者本人， 转载请联系作者并注明出处：http://www.betaflare.com/web/1744646972a2617452.html， 本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如发现本站有涉嫌抄袭侵权/违法违规的内容，一经查实，本站将立刻删除。