# OpenDataLab Python SDK
[](https://pepy.tech/project/opendatalab)
[](https://pypi.org/project/opendatalab/)
[](https://pypi.org/project/opendatalab/)
---
**IMPORTANT**: OpenDataLab SDK WIP, not ensure the necessary compatibility of OpenAPI and SDK. As a result, please use the SDK with the **latest** version.
---
OpenDataLab Python SDK is a python library to access [Opendatalab](https://opendatalab.org.cn/)
and use open datasets.
It provides:
- A pythonic way to access opendatalab resources.
- A convenient CLI tool `odl` to access open datasets.
## Installation
```console
$ pip3 install opendatalab
```
## Usage:
An **account** is needed to access to opendatalab platform.
Please visit [offical websit](https://opendatalab.org.cn/register) to get the account username and password first.
### Help
Show cmd help
```cmd
$ odl -h
$ odl --help
Usage: odl [OPTIONS] COMMAND [ARGS]...
You can use `odl <command>` to access open datasets.
Options:
--version Show the version and exit.
-h, --help Show this message and exit.
Commands:
get Get(Download) dataset files into local path.
info Print dataset info.
login Login opendatalab with account.
logout Logout opendatalab account.
ls List files of the dataset.
search Search dataset info.
version Show opendatalab version.
```
### Version
```cmd
$ odl version
odl version, current: 0.0.6, svc: 1.8
```
### Login
Login with opendatalab username and password. If you haven't an opendatalab account，please register with link: https://opendatalab.org.cn/
```cmd
$ odl login
Username []: someone@example.com
Password []:
Login successfully as someone@example.com
or
$ odl login -u someone@example.com
Password[]:
```
### Logout
Logout current opendatalab account
```cmd
$ odl logout
Do you want to logout? [y/N]: y
someone@example.com logout
```
### List Dataset Files
List dataset files, support prefix of sub_directory
```cmd
# list all dataset files
$ odl ls MNIST
total: 4, size: 11.1M
+----------------------------+--------------+
| File Name | Size |
+----------------------------+--------------+
| train-labels-idx1-ubyte.gz | 28.2K |
+----------------------------+--------------+
| train-images-idx3-ubyte.gz | 9.5M |
+----------------------------+--------------+
| t10k-labels-idx1-ubyte.gz | 4.4K |
+----------------------------+--------------+
| t10k-images-idx3-ubyte.gz | 1.6M |
+----------------------------+--------------+ 1.6M
# list sub directory files
$ odl ls MNIST/t10k
total: 2, size: 1.6M
+---------------------------+--------------+
| File Name | Size |
+---------------------------+--------------+
| t10k-labels-idx1-ubyte.gz | 4.4K |
+---------------------------+--------------+
| t10k-images-idx3-ubyte.gz | 1.6M |
+---------------------------+--------------+
```
```cmd
# download dataset files into local
# get all files of dataset
$ odl get MNIST
# get partial files of dataset
$ odl get MNIST/t10k
```
## Python Develop Sample
```python
import json
from opendatalab.__version__ import __url__
from opendatalab.cli.get import implement_get
from opendatalab.cli.info import implement_info
from opendatalab.cli.login import implement_login
from opendatalab.cli.ls import implement_ls
from opendatalab.cli.search import implement_search
from opendatalab.cli.utility import ContextInfo
if __name__ == '__main__':
"""
ContextInfo: default
please use shell login first, use: opendatalab login
"""
ctx = ContextInfo(__url__, "")
client = ctx.get_client()
odl_api = client.get_api()
# 0. login with account
# account = "xxxxx" # your username
# pw = "xxxxx" # your password
# print(f'*****'*8)
# implement_login(ctx, account, pw)
# 1. search demo
res_list = odl_api.search_dataset("coco")
for index, res in enumerate(res_list):
print(f"index: {index}, result: {res['name']}")
# implement_search("coco")
print(f'*****'*8)
# 2. list demo
implement_ls(ctx, 'TAO')
print(f'*****' * 8)
# 3. read file online demo
dataset = client.get_dataset('FB15k')
with dataset.get('meta/info.json', False) as fd:
content = json.load(fd)
print(f"{content}")
print(f'*****'*8)
# 4. get dataset info
implement_info(ctx, 'FB15k')
# 5. download
# get all files of dataset
# implement_get(ctx, "MNIST", 4, 0)
# get partial files of dataset
implement_get(ctx, "GOT-10k/data/test_data.zip", 4, 0) # 139, zip 1.16G GOT-10k
print(f'*****' * 5)
```
## Documentation
More information can be found on the [documentation site](https://opendatalab.org.cn/docs)
Raw data
{
"_id": null,
"home_page": "https://github.com/opendatalab/opendatalab-python-sdk",
"name": "opendatalab",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.7",
"maintainer_email": "",
"keywords": "opendatalab,dataset,test",
"author": "OpenDataLab",
"author_email": "OpenDataLab@pjlab.org.cn",
"download_url": "https://files.pythonhosted.org/packages/74/9f/25bfae72e3d10040f6ba80e2b0b9688c9477528b2aed1fe871847f48e479/opendatalab-0.0.10.tar.gz",
"platform": null,
"description": "# OpenDataLab Python SDK\n\n\n[](https://pepy.tech/project/opendatalab)\n[](https://pypi.org/project/opendatalab/)\n[](https://pypi.org/project/opendatalab/)\n\n---\n\n**IMPORTANT**: OpenDataLab SDK WIP, not ensure the necessary compatibility of OpenAPI and SDK. As a result, please use the SDK with the **latest** version. \n\n---\n\nOpenDataLab Python SDK is a python library to access [Opendatalab](https://opendatalab.org.cn/)\nand use open datasets. \nIt provides:\n\n- A pythonic way to access opendatalab resources.\n- A convenient CLI tool `odl` to access open datasets.\n\n## Installation\n\n```console\n$ pip3 install opendatalab\n```\n\n## Usage:\n\nAn **account** is needed to access to opendatalab platform.\nPlease visit [offical websit](https://opendatalab.org.cn/register) to get the account username and password first.\n\n### Help\nShow cmd help\n```cmd\n$ odl -h\n$ odl --help\n\nUsage: odl [OPTIONS] COMMAND [ARGS]...\n\n You can use `odl <command>` to access open datasets.\n\nOptions:\n --version Show the version and exit.\n -h, --help Show this message and exit.\n\nCommands:\n get Get(Download) dataset files into local path.\n info Print dataset info.\n login Login opendatalab with account.\n logout Logout opendatalab account.\n ls List files of the dataset.\n search Search dataset info.\n version Show opendatalab version.\n```\n\n### Version\n```cmd\n$ odl version\nodl version, current: 0.0.6, svc: 1.8\n```\n\n### Login\nLogin with opendatalab username and password. If you haven't an opendatalab account\uff0cplease register with link: https://opendatalab.org.cn/\n\n```cmd\n$ odl login\nUsername []: someone@example.com\nPassword []: \nLogin successfully as someone@example.com\nor\n$ odl login -u someone@example.com\nPassword[]:\n```\n\n### Logout\nLogout current opendatalab account \n```cmd\n$ odl logout\nDo you want to logout? [y/N]: y\nsomeone@example.com logout\n```\n\n\n### List Dataset Files\nList dataset files, support prefix of sub_directory\n```cmd\n# list all dataset files \n$ odl ls MNIST\ntotal: 4, size: 11.1M\n+----------------------------+--------------+\n| File Name | Size |\n+----------------------------+--------------+\n| train-labels-idx1-ubyte.gz | 28.2K |\n+----------------------------+--------------+\n| train-images-idx3-ubyte.gz | 9.5M |\n+----------------------------+--------------+\n| t10k-labels-idx1-ubyte.gz | 4.4K |\n+----------------------------+--------------+\n| t10k-images-idx3-ubyte.gz | 1.6M |\n+----------------------------+--------------+ \t1.6M\n\n# list sub directory files\n$ odl ls MNIST/t10k\ntotal: 2, size: 1.6M\n+---------------------------+--------------+\n| File Name | Size |\n+---------------------------+--------------+\n| t10k-labels-idx1-ubyte.gz | 4.4K |\n+---------------------------+--------------+\n| t10k-images-idx3-ubyte.gz | 1.6M |\n+---------------------------+--------------+\n```\n\n```cmd\n# download dataset files into local \n# get all files of dataset \n$ odl get MNIST \n\n# get partial files of dataset \n$ odl get MNIST/t10k \n```\n\n## Python Develop Sample\n```python\nimport json\nfrom opendatalab.__version__ import __url__\nfrom opendatalab.cli.get import implement_get\nfrom opendatalab.cli.info import implement_info\nfrom opendatalab.cli.login import implement_login\nfrom opendatalab.cli.ls import implement_ls\nfrom opendatalab.cli.search import implement_search\nfrom opendatalab.cli.utility import ContextInfo\n\nif __name__ == '__main__':\n \"\"\"\n ContextInfo: default\n please use shell login first, use: opendatalab login\n \"\"\"\n ctx = ContextInfo(__url__, \"\")\n client = ctx.get_client()\n odl_api = client.get_api()\n\n # 0. login with account\n # account = \"xxxxx\" # your username\n # pw = \"xxxxx\" # your password\n # print(f'*****'*8)\n # implement_login(ctx, account, pw)\n\n # 1. search demo \n res_list = odl_api.search_dataset(\"coco\")\n for index, res in enumerate(res_list):\n print(f\"index: {index}, result: {res['name']}\")\n\n # implement_search(\"coco\")\n print(f'*****'*8)\n\n # 2. list demo\n implement_ls(ctx, 'TAO')\n print(f'*****' * 8)\n\n # 3. read file online demo\n dataset = client.get_dataset('FB15k')\n with dataset.get('meta/info.json', False) as fd:\n content = json.load(fd)\n print(f\"{content}\")\n print(f'*****'*8)\n\n # 4. get dataset info\n implement_info(ctx, 'FB15k')\n\n # 5. download\n # get all files of dataset\n # implement_get(ctx, \"MNIST\", 4, 0)\n\n # get partial files of dataset\n implement_get(ctx, \"GOT-10k/data/test_data.zip\", 4, 0) # 139, zip 1.16G GOT-10k\n print(f'*****' * 5)\n```\n\n## Documentation\nMore information can be found on the [documentation site](https://opendatalab.org.cn/docs)\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "OpenDataLab Python SDK",
"version": "0.0.10",
"project_urls": {
"Bug Tracker": "https://github.com/opendatalab/opendatalab-python-sdk/issues",
"Homepage": "https://github.com/opendatalab/opendatalab-python-sdk"
},
"split_keywords": [
"opendatalab",
"dataset",
"test"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "108228fa3a91b7c4852fbad9ad32c7b49e4b1e212ab7ccf7296736da0935070d",
"md5": "0ac52fccb867193605e9d389fe075815",
"sha256": "b6a317785b7db418739933d4af6d981a0e45f6cf20a3e113bef63ed9b4488251"
},
"downloads": -1,
"filename": "opendatalab-0.0.10-py3-none-any.whl",
"has_sig": false,
"md5_digest": "0ac52fccb867193605e9d389fe075815",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.7",
"size": 29506,
"upload_time": "2023-08-02T07:29:58",
"upload_time_iso_8601": "2023-08-02T07:29:58.629288Z",
"url": "https://files.pythonhosted.org/packages/10/82/28fa3a91b7c4852fbad9ad32c7b49e4b1e212ab7ccf7296736da0935070d/opendatalab-0.0.10-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "749f25bfae72e3d10040f6ba80e2b0b9688c9477528b2aed1fe871847f48e479",
"md5": "65c8db6cf3099503a46d5dbe1b00a5db",
"sha256": "9b1382f974bd76a961747dc33308fce5b024d337c02cf3acb728c64952ca9aaf"
},
"downloads": -1,
"filename": "opendatalab-0.0.10.tar.gz",
"has_sig": false,
"md5_digest": "65c8db6cf3099503a46d5dbe1b00a5db",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.7",
"size": 23615,
"upload_time": "2023-08-02T07:30:08",
"upload_time_iso_8601": "2023-08-02T07:30:08.352903Z",
"url": "https://files.pythonhosted.org/packages/74/9f/25bfae72e3d10040f6ba80e2b0b9688c9477528b2aed1fe871847f48e479/opendatalab-0.0.10.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-08-02 07:30:08",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "opendatalab",
"github_project": "opendatalab-python-sdk",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"lcname": "opendatalab"
}
JSON
