Machine learning prediction serving
Project description
ServeIt deploys your trained models to a RESTful API for prediction serving. Current features include:
Model prediction serving
Supplementary information endpoint creation
Input validation and exception handling
Configurable request and response logging (work in progress)
Installation: Python 2.7 and Python 3.6
PyPi: pip install serveit
source: python setup.py
Supported libraries
Scikit-Learn
Fit Scikit-Learn model:
from sklearn.datasets import load_iris
from sklearn.linear_model import LogisticRegression
# fit a model on the Iris dataset
data = load_iris()
reg = LogisticRegression()
reg.fit(data.data, data.target)
Serve your trained model:
from serveit.sklearn_server import SklearnServer
# initialize server
sklearn_server = SklearnServer(reg, reg.predict)
# add (optional) informational endpoints
sklearn_server.create_model_info_endpoint()
sklearn_server.create_info_endpoint('features', data.feature_names)
sklearn_server.create_info_endpoint('target_labels', data.target_names.tolist())
# start serving predictions from API
sklearn_server.serve()
Then try out your new API:
curl -XPOST 'localhost:5000/predictions'\
-H "Content-Type: application/json"\
-d "[[5.6, 2.9, 3.6, 1.3], [4.4, 2.9, 1.4, 0.2], [5.5, 2.4, 3.8, 1.1], [5.0, 3.4, 1.5, 0.2], [5.7, 2.5, 5.0, 2.0]]"
# [1, 0, 1, 0, 2]
curl -XGET 'localhost:5000/info/model'
# {"penalty": "l2", "tol": 0.0001, "C": 1.0, "classes_": [0, 1, 2], "coef_": [[0.4150, 1.4613, -2.2621, -1.0291], ...], ...}
curl -XGET 'localhost:5000/info/features'
# ["sepal length (cm)", "sepal width (cm)", "petal length (cm)", "petal width (cm)"]
curl -XGET 'localhost:5000/info/target_labels'
# ["setosa", "versicolor", "virginica"]
Coming soon:
TensorFlow
Keras
PyTorch
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
ServeIt-0.0.2a6.tar.gz
(8.5 kB
view hashes)
Built Distribution
Close
Hashes for ServeIt-0.0.2a6-py2.py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 9f386f051f0da0912b837b3f460f6594379b6d698d9a45a49806a284a6b036a5 |
|
MD5 | 9146b6bd59596eb9b3e6d640d7566475 |
|
BLAKE2b-256 | 9f8cc0161f5b4652b3a4d8151c08985048ddbea9fa73b992b574552eca2b5936 |