# Jai SDK - Trust your data
[](https://pypi.org/project/jai-sdk/)
[](https://img.shields.io/badge/python-3.7%20%7C%203.8-blue)
[](https://jai-sdk.readthedocs.io/en/latest/?badge=latest)
[](https://codecov.io/gh/jquant/jai-sdk)
[](https://github.com/jquant/jai-sdk/blob/main/LICENSE)
[](https://github.com/google/yapf)
[](https://pepy.tech/project/jai-sdk)
# Installation
The source code is currently hosted on GitHub at: [https://github.com/jquant/jai-sdk](https://github.com/jquant/jai-sdk)
The latest version of JAI-SDK can be installed from `pip`:
```sh
pip install jai-sdk --user
```
Nowadays, JAI supports python 3.7+. For more information, here is our [documentation](https://jai-sdk.readthedocs.io/en/latest/).
# Getting your auth key
JAI requires an auth key to organize and secure collections.
You can quickly generate your free-forever auth-key by running the command below:
```python
from jai import get_auth_key
get_auth_key(email='email@mail.com', firstName='Jai', lastName='Z')
```
> **_ATTENTION:_** Your auth key will be sent to your e-mail, so please make sure to use a valid address and check your spam folder.
# How does it work?
With JAI, you can train models in the cloud and run inference on your trained models. Besides, you can achieve all your models through a REST API endpoint.
First, you can set your auth key into an environment variable or use a :file:`.env` file or :file:`.ini` file.
Please check the section [How to configure your auth key](https://jai-sdk.readthedocs.io/en/latest/source/overview/set_authentication.html>) for more information.
Bellow an example of the content of the :file:`.env` file:
```text
JAI_AUTH="xXxxxXXxXXxXXxXXxXXxXXxXXxxx"
```
In the below example, we'll show how to train a simple supervised model (regression) using the California housing dataset, run a prediction from this model, and call this prediction directly from the REST API.
```python
import pandas as pd
from jai import Jai
from sklearn.datasets import fetch_california_housing
# Load dataset
data, labels = fetch_california_housing(as_frame=True, return_X_y=True)
model_data = pd.concat([data, labels], axis=1)
# Instanciating JAI class
j = Jai()
# Send data to JAI for feature extraction
j.fit(
name='california_supervised', # JAI collection name
data=model_data, # Data to be processed
db_type='Supervised', # Your training type ('Supervised', 'SelfSupervised' etc)
verbose=2,
hyperparams={
'learning_rate': 3e-4,
'pretraining_ratio': 0.8
},
label={
'task': 'regression',
'label_name': 'MedHouseVal'
},
overwrite=True)
# Run prediction
j.predict(name='california_supervised', data=data)
```
In this example, you could train a supervised model with the California housing dataset and run a prediction with some data.
JAI supports many other training models, like self-supervised model training.
Besides, it also can train on different data types, like text and images.
You can find a complete list of the model types supported by JAI on [The Fit Method](https://jai-sdk.readthedocs.io/en/latest/source/using_jai/fit.html).
# Read our documentation
For more information, here is our [documentation](https://jai-sdk.readthedocs.io/en/latest/).
Raw data
{
"_id": null,
"home_page": "https://github.com/jquant/jai-sdk",
"name": "jai-sdk",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.9",
"maintainer_email": "",
"keywords": "",
"author": "JQuant",
"author_email": "jedis@jquant.com.br",
"download_url": "https://files.pythonhosted.org/packages/4f/31/f8365244b9e11d871dacd1b27752a3a6b7e918ccaf0b3f253d8cfc169620/jai-sdk-0.25.0.tar.gz",
"platform": null,
"description": "# Jai SDK - Trust your data\n\n[](https://pypi.org/project/jai-sdk/)\n[](https://img.shields.io/badge/python-3.7%20%7C%203.8-blue)\n[](https://jai-sdk.readthedocs.io/en/latest/?badge=latest)\n[](https://codecov.io/gh/jquant/jai-sdk)\n[](https://github.com/jquant/jai-sdk/blob/main/LICENSE)\n[](https://github.com/google/yapf)\n[](https://pepy.tech/project/jai-sdk)\n\n# Installation\n\nThe source code is currently hosted on GitHub at: [https://github.com/jquant/jai-sdk](https://github.com/jquant/jai-sdk)\n\nThe latest version of JAI-SDK can be installed from `pip`:\n\n```sh\npip install jai-sdk --user\n```\n\nNowadays, JAI supports python 3.7+. For more information, here is our [documentation](https://jai-sdk.readthedocs.io/en/latest/).\n\n# Getting your auth key\n\nJAI requires an auth key to organize and secure collections.\nYou can quickly generate your free-forever auth-key by running the command below:\n\n```python\nfrom jai import get_auth_key\nget_auth_key(email='email@mail.com', firstName='Jai', lastName='Z')\n```\n\n> **_ATTENTION:_** Your auth key will be sent to your e-mail, so please make sure to use a valid address and check your spam folder.\n\n# How does it work?\n\nWith JAI, you can train models in the cloud and run inference on your trained models. Besides, you can achieve all your models through a REST API endpoint.\n\nFirst, you can set your auth key into an environment variable or use a :file:`.env` file or :file:`.ini` file.\nPlease check the section [How to configure your auth key](https://jai-sdk.readthedocs.io/en/latest/source/overview/set_authentication.html>) for more information.\n\nBellow an example of the content of the :file:`.env` file:\n\n```text\nJAI_AUTH=\"xXxxxXXxXXxXXxXXxXXxXXxXXxxx\"\n```\n\nIn the below example, we'll show how to train a simple supervised model (regression) using the California housing dataset, run a prediction from this model, and call this prediction directly from the REST API.\n\n```python\nimport pandas as pd\nfrom jai import Jai\nfrom sklearn.datasets import fetch_california_housing\n\n# Load dataset\ndata, labels = fetch_california_housing(as_frame=True, return_X_y=True)\nmodel_data = pd.concat([data, labels], axis=1)\n\n# Instanciating JAI class\nj = Jai()\n\n# Send data to JAI for feature extraction\nj.fit(\n name='california_supervised', # JAI collection name\n data=model_data, # Data to be processed\n db_type='Supervised', # Your training type ('Supervised', 'SelfSupervised' etc)\n verbose=2,\n hyperparams={\n 'learning_rate': 3e-4,\n 'pretraining_ratio': 0.8\n },\n label={\n 'task': 'regression',\n 'label_name': 'MedHouseVal'\n },\n overwrite=True)\n# Run prediction\nj.predict(name='california_supervised', data=data)\n```\n\nIn this example, you could train a supervised model with the California housing dataset and run a prediction with some data.\n\nJAI supports many other training models, like self-supervised model training.\nBesides, it also can train on different data types, like text and images.\nYou can find a complete list of the model types supported by JAI on [The Fit Method](https://jai-sdk.readthedocs.io/en/latest/source/using_jai/fit.html).\n\n# Read our documentation\n\nFor more information, here is our [documentation](https://jai-sdk.readthedocs.io/en/latest/).\n\n\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "JAI - Trust your data",
"version": "0.25.0",
"project_urls": {
"Homepage": "https://github.com/jquant/jai-sdk"
},
"split_keywords": [],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "fdf558134614a4b6620686f45270da3181a9bf5086c8179e9b3be8a37fca5ee0",
"md5": "15b326572613b41552ba3a8e6f47f5fd",
"sha256": "b948ecf837688485db38e94d2ea8db8aeab0a92c99a875bf1b70d559a922e47c"
},
"downloads": -1,
"filename": "jai_sdk-0.25.0-py3-none-any.whl",
"has_sig": false,
"md5_digest": "15b326572613b41552ba3a8e6f47f5fd",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.9",
"size": 64846,
"upload_time": "2023-10-05T05:00:40",
"upload_time_iso_8601": "2023-10-05T05:00:40.465983Z",
"url": "https://files.pythonhosted.org/packages/fd/f5/58134614a4b6620686f45270da3181a9bf5086c8179e9b3be8a37fca5ee0/jai_sdk-0.25.0-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "4f31f8365244b9e11d871dacd1b27752a3a6b7e918ccaf0b3f253d8cfc169620",
"md5": "9438cdf4076f5acc2e50dbfa7f91db92",
"sha256": "aba8fdc731133b240cc2b631bb4d96a210373fe160f9e04887c55bcd5746711b"
},
"downloads": -1,
"filename": "jai-sdk-0.25.0.tar.gz",
"has_sig": false,
"md5_digest": "9438cdf4076f5acc2e50dbfa7f91db92",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.9",
"size": 53326,
"upload_time": "2023-10-05T05:00:42",
"upload_time_iso_8601": "2023-10-05T05:00:42.343090Z",
"url": "https://files.pythonhosted.org/packages/4f/31/f8365244b9e11d871dacd1b27752a3a6b7e918ccaf0b3f253d8cfc169620/jai-sdk-0.25.0.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-10-05 05:00:42",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "jquant",
"github_project": "jai-sdk",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"requirements": [
{
"name": "numpy",
"specs": [
[
">=",
"1.21.0"
]
]
},
{
"name": "pandas",
"specs": [
[
">=",
"1.3.0"
]
]
},
{
"name": "tqdm",
"specs": [
[
">=",
"4.61.2"
]
]
},
{
"name": "pillow",
"specs": [
[
">=",
"8.3.2"
]
]
},
{
"name": "psutil",
"specs": [
[
">=",
"5.9.0"
]
]
},
{
"name": "pydantic",
"specs": [
[
">=",
"2.0.0"
]
]
},
{
"name": "python-decouple",
"specs": [
[
">=",
"3.6"
]
]
},
{
"name": "matplotlib",
"specs": [
[
">=",
"3.4.2"
]
]
},
{
"name": "requests",
"specs": []
},
{
"name": "scikit-learn",
"specs": [
[
">=",
"0.24.2"
]
]
}
],
"lcname": "jai-sdk"
}
JSON
