from torchvision import datasets
时间: 2023-11-14 22:05:55 浏览: 38
The torchvision module provides access to popular datasets for computer vision tasks. Some of the datasets available in torchvision are:
1. MNIST: This is a dataset of handwritten digits and is widely used for image classification tasks.
2. CIFAR-10: This dataset consists of 60,000 32x32 RGB images in 10 classes, with 6,000 images per class.
3. ImageNet: This is a large-scale image recognition dataset with over 1 million images and 1,000 classes.
4. COCO: This dataset contains images of complex everyday scenes and is used for object detection, segmentation, and captioning tasks.
5. Fashion-MNIST: This is a dataset of fashion items such as clothes, shoes, and bags, and is used for image classification tasks.
These datasets can be easily downloaded and loaded into your Python code using the `datasets` module from torchvision.