public dataset

View dataset

tip

If the platform does not have the dataset you need, please leave a message in the forum dataset section, the technology will be updated within 24 hours .

The platform provides some common datasets, you can view the list of datasets in Instances and Data - public dataset . The dataset name search is a fuzzy search, the dataset name and the dataset folder have the same name, without whitespace characters. When searching, you can enter a word of the dataset name or dataset abbreviation to find it.

use dataset

Public datasets can be downloaded using the oss command line tool. Use the -public parameter to browse and download files in public datasets, and use the -ent-pub parameter for private cloud users to browse and upload and download enterprise datasets.

# Log in to the Hengyuan cloud account, use the account name and password of Hengyuan cloud, the account name is the mobile phone number
# If it is a non-mainland China mobile phone number, you need to add the area code with +
 oss login
Username: 139********
Password:**********
139******** login successfully!

# If you are a proprietary cloud user, you need to add the -cloud=ent parameter to log in to the proprietary cloud
# The format of the username is proprietary cloud username@enterprise ID
 ./oss login -cloud=ent
Username: username@panda
Password:**********
username@panda login successfully!

# View the directory of public datasets (proprietary cloud users need to replace -public with -ent-pub to view enterprise datasets)
 oss ls -public -s -d oss://

# View files of a public dataset (proprietary cloud users need to replace -public with -ent-pub to view enterprise datasets)
 oss ls -public -s oss://MNIST

# Download the MNIST directory in the public dataset to the /hy-tmp directory (proprietary cloud users need to replace -public with -ent-pub to view enterprise datasets)
 oss cp -r -f -public oss://MNIST /hy-tmp
 cd /hy-tmp
/hy-tmp# ls
MNIST

# Download a file in the public dataset to the /hy-tmp directory (proprietary cloud users need to replace -public with -ent-pub to view enterprise datasets)
 oss cp -public oss://MNIST/t10k-images-idx3-ubyte.gz /hy-tmp

# Proprietary cloud users can upload data to the enterprise data set
 oss cp -ent-pub MNIST.tar.gz oss://

Or enter the instance terminal through an SSH client or JupyterLab, and the public dataset can be found in the /hy-public folder. When using a dataset, first copy the dataset to hy-tmp or /hy-nas (only for shared storage models), and then decompress it.

 cd /hy-public/
/hy-public# ls
A2D2 ADE20K Argoverse BSTLD CCPD CIFAR COCO CUHK CWT MNIST...

/hy-public# cp -r MNIST /hy-tmp
/hy-public# cd /hy-tmp
/hy-tmp# ls
MNIST