skl2neonsvm 0.1.0

Convert scikit-learn SVM based models to neonsvm format

skl2neonsvm

Library for converting scikit-learn SVM-based classification and regression models to neonsvm YAML format for optimized inference on ARM devices using C++ library neonsvm.

Installation

From PyPi

pip install skl2neonsvm

From repository

pip install git+https://github.com/R3dKar/sklearn-neonsvm

Usage

# Training model
from sklearn.datasets import make_classification
from sklearn.svm import SVC

X, y = make_classification(random_state=42)
model = SVC().fit(X, y)

# Converting and exporting
from skl2neonsvm import to_neonsvm

data = to_neonsvm(model)
with open("model.yaml", "w") as f:
    f.write(data)

Supported models

All models of submodule sklearn.svm are supported.

Supported classification models:

• SVC
• NuSVC
• LinearSVC
• OneClassSVM

Supported regression models:

• SVR
• NuSVR
• LinearSVR