<p align="center">
<a href="https://www.union.ai/unionml">
<img src="https://raw.githubusercontent.com/unionai-oss/unionml/main/docs/source/_static/images/union-logo.svg" alt="Union.ai Logo" width="100">
</a>
</p>
<h1 align="center">UnionML</h1>
<p align="center">
<strong>The easiest way to build and deploy machine learning microservices</strong>
</p>
---
<br>
[](https://pypi.org/project/unionml/)
[](https://pypi.org/project/unionml/)
[](https://unionml.readthedocs.io/en/latest/?badge=latest)
[](https://github.com/unionai-oss/unionml/actions/workflows/build.yml)
[](https://pypistats.org/packages/unionml)
[](https://github.com/orgs/unionai-oss/projects/1/views/4)
[](https://app.addevent.com/event/tj14110550/)
<br>
**UnionML** is an open source MLOps framework that aims to reduce the boilerplate and friction
that comes with building models and deploying them to production.
You can create **UnionML Apps** by defining a few core methods that are automatically bundled
into ML microservices, starting with model training and offline and online prediction.
Built on top of [Flyte](https://docs.flyte.org/en/latest/), UnionML provides a high-level
interface for productionizing your ML models so that you can focus on curating a better dataset
and improving your models.
To learn more, check out the 📖 [Documentation](https://unionml.readthedocs.io).
## Installing
Install using conda:
```bash
conda install -c conda-forge unionml
```
Install using pip:
```bash
pip install unionml
```
## A Simple Example
Create a `Dataset` and `Model`, which together form a **UnionML App**:
```python
from unionml import Dataset, Model
from sklearn.linear_model import LogisticRegression
dataset = Dataset(name="digits_dataset", test_size=0.2, shuffle=True, targets=["target"])
model = Model(name="digits_classifier", init=LogisticRegression, dataset=dataset)
```
Define `Dataset` and `Model` methods for training a hand-written digits classifier:
```python
from typing import List
import pandas as pd
from sklearn.datasets import load_digits
from sklearn.metrics import accuracy_score
@dataset.reader
def reader() -> pd.DataFrame:
return load_digits(as_frame=True).frame
@model.trainer
def trainer(
estimator: LogisticRegression,
features: pd.DataFrame,
target: pd.DataFrame,
) -> LogisticRegression:
return estimator.fit(features, target.squeeze())
@model.predictor
def predictor(
estimator: LogisticRegression,
features: pd.DataFrame
) -> List[float]:
return [float(x) for x in estimator.predict(features)]
@model.evaluator
def evaluator(
estimator: LogisticRegression,
features: pd.DataFrame,
target: pd.DataFrame
) -> float:
return float(accuracy_score(target.squeeze(), predictor(estimator, features)))
```
And that's all ⭐️!
By defining these four methods, you've created a minimal **UnionML App** that you can:
- [Execute locally](https://unionml.readthedocs.io/en/latest/index.html#train-and-predict-locally) to debug and iterate on your code.
- [Serve Seamlessly with FastAPI](https://unionml.readthedocs.io/en/latest/index.html#serve-seamlessly-with-fastapi) for online prediction.
- [Deploy on a Flyte Cluster](https://unionml.readthedocs.io/en/latest/deploying.html) to scale your model training and schedule offline prediction jobs.
## Contributing
All contributions are welcome 🤝 ! Check out the [contribution guide](https://unionml.readthedocs.io/en/latest/contributing.html) to learn more about how to contribute.
### Gitpod
[](https://gitpod.io/#https://github.com/unionai-oss/unionml)
Raw data
{
"_id": null,
"home_page": "https://github.com/unionai-oss/unionml/",
"name": "unionml",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.7",
"maintainer_email": "",
"keywords": "machine-learning,artificial-intelligence,microservices",
"author": "unionai-oss",
"author_email": "info@union.ai",
"download_url": "https://files.pythonhosted.org/packages/c9/1d/358c7d478d1a30446961126f609b49b2401d118637c57dbb029245c79fac/unionml-0.2.1.tar.gz",
"platform": "any",
"description": "<p align=\"center\">\n <a href=\"https://www.union.ai/unionml\">\n <img src=\"https://raw.githubusercontent.com/unionai-oss/unionml/main/docs/source/_static/images/union-logo.svg\" alt=\"Union.ai Logo\" width=\"100\">\n </a>\n</p>\n\n<h1 align=\"center\">UnionML</h1>\n\n<p align=\"center\">\n <strong>The easiest way to build and deploy machine learning microservices</strong>\n</p>\n\n---\n\n<br>\n\n[](https://pypi.org/project/unionml/)\n[](https://pypi.org/project/unionml/)\n[](https://unionml.readthedocs.io/en/latest/?badge=latest)\n[](https://github.com/unionai-oss/unionml/actions/workflows/build.yml)\n[](https://pypistats.org/packages/unionml)\n[](https://github.com/orgs/unionai-oss/projects/1/views/4)\n[](https://app.addevent.com/event/tj14110550/)\n\n<br>\n\n\n**UnionML** is an open source MLOps framework that aims to reduce the boilerplate and friction\nthat comes with building models and deploying them to production.\n\nYou can create **UnionML Apps** by defining a few core methods that are automatically bundled\ninto ML microservices, starting with model training and offline and online prediction.\n\nBuilt on top of [Flyte](https://docs.flyte.org/en/latest/), UnionML provides a high-level\ninterface for productionizing your ML models so that you can focus on curating a better dataset\nand improving your models.\n\nTo learn more, check out the \ud83d\udcd6 [Documentation](https://unionml.readthedocs.io).\n\n## Installing\n\nInstall using conda:\n\n```bash\nconda install -c conda-forge unionml\n```\n\nInstall using pip:\n\n```bash\npip install unionml\n```\n\n## A Simple Example\n\nCreate a `Dataset` and `Model`, which together form a **UnionML App**:\n\n```python\nfrom unionml import Dataset, Model\n\nfrom sklearn.linear_model import LogisticRegression\n\ndataset = Dataset(name=\"digits_dataset\", test_size=0.2, shuffle=True, targets=[\"target\"])\nmodel = Model(name=\"digits_classifier\", init=LogisticRegression, dataset=dataset)\n```\n\nDefine `Dataset` and `Model` methods for training a hand-written digits classifier:\n\n```python\nfrom typing import List\n\nimport pandas as pd\nfrom sklearn.datasets import load_digits\nfrom sklearn.metrics import accuracy_score\n\n@dataset.reader\ndef reader() -> pd.DataFrame:\n return load_digits(as_frame=True).frame\n\n@model.trainer\ndef trainer(\n estimator: LogisticRegression,\n features: pd.DataFrame,\n target: pd.DataFrame,\n) -> LogisticRegression:\n return estimator.fit(features, target.squeeze())\n\n@model.predictor\ndef predictor(\n estimator: LogisticRegression,\n features: pd.DataFrame\n) -> List[float]:\n return [float(x) for x in estimator.predict(features)]\n\n@model.evaluator\ndef evaluator(\n estimator: LogisticRegression,\n features: pd.DataFrame,\n target: pd.DataFrame\n) -> float:\n return float(accuracy_score(target.squeeze(), predictor(estimator, features)))\n```\n\nAnd that's all \u2b50\ufe0f!\n\nBy defining these four methods, you've created a minimal **UnionML App** that you can:\n\n- [Execute locally](https://unionml.readthedocs.io/en/latest/index.html#train-and-predict-locally) to debug and iterate on your code.\n- [Serve Seamlessly with FastAPI](https://unionml.readthedocs.io/en/latest/index.html#serve-seamlessly-with-fastapi) for online prediction.\n- [Deploy on a Flyte Cluster](https://unionml.readthedocs.io/en/latest/deploying.html) to scale your model training and schedule offline prediction jobs.\n\n## Contributing\n\nAll contributions are welcome \ud83e\udd1d ! Check out the [contribution guide](https://unionml.readthedocs.io/en/latest/contributing.html) to learn more about how to contribute.\n\n### Gitpod\n\n[](https://gitpod.io/#https://github.com/unionai-oss/unionml)\n",
"bugtrack_url": null,
"license": "Apache",
"summary": "The easiest way to build and deploy machine learning microservices.",
"version": "0.2.1",
"split_keywords": [
"machine-learning",
"artificial-intelligence",
"microservices"
],
"urls": [
{
"comment_text": "",
"digests": {
"md5": "c3112d94c654605f97023dfc53cd133b",
"sha256": "ae891baf744b2e4c2aa8445d92121158490a07b868b3e1c7f1033424e14e75bb"
},
"downloads": -1,
"filename": "unionml-0.2.1-py3-none-any.whl",
"has_sig": false,
"md5_digest": "c3112d94c654605f97023dfc53cd133b",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.7",
"size": 87745,
"upload_time": "2022-12-20T18:57:40",
"upload_time_iso_8601": "2022-12-20T18:57:40.297351Z",
"url": "https://files.pythonhosted.org/packages/2f/fe/31ce6b74e41e6335233e8b5acafef2a3d200e13a92513006a7c973e54b74/unionml-0.2.1-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"md5": "6a5abe3c1c94ebec750ec416c2d5d65c",
"sha256": "c237a74a0599a98171a49e17ccdb138f6678b8f227d16f0d7eff154abea45d12"
},
"downloads": -1,
"filename": "unionml-0.2.1.tar.gz",
"has_sig": false,
"md5_digest": "6a5abe3c1c94ebec750ec416c2d5d65c",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.7",
"size": 61390,
"upload_time": "2022-12-20T18:57:42",
"upload_time_iso_8601": "2022-12-20T18:57:42.439234Z",
"url": "https://files.pythonhosted.org/packages/c9/1d/358c7d478d1a30446961126f609b49b2401d118637c57dbb029245c79fac/unionml-0.2.1.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2022-12-20 18:57:42",
"github": true,
"gitlab": false,
"bitbucket": false,
"github_user": "unionai-oss",
"github_project": "unionml",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"requirements": [
{
"name": "click",
"specs": []
},
{
"name": "cookiecutter",
"specs": []
},
{
"name": "dataclasses_json",
"specs": []
},
{
"name": "docker",
"specs": []
},
{
"name": "fastapi",
"specs": []
},
{
"name": "flytekit",
"specs": [
[
">=",
"1.1.0"
]
]
},
{
"name": "gitpython",
"specs": []
},
{
"name": "joblib",
"specs": []
},
{
"name": "mangum",
"specs": []
},
{
"name": "numpy",
"specs": []
},
{
"name": "pandas",
"specs": []
},
{
"name": "pydantic",
"specs": []
},
{
"name": "rich",
"specs": [
[
">=",
"11.2.0"
]
]
},
{
"name": "scikit-learn",
"specs": []
},
{
"name": "typer",
"specs": []
},
{
"name": "typing-extensions",
"specs": [
[
">=",
"4.3.0"
]
]
},
{
"name": "uvicorn",
"specs": []
},
{
"name": "importlib-metadata",
"specs": []
}
],
"lcname": "unionml"
}
JSON
