mlprodict was initially started to help implementing converters to ONNX. The main features is a python runtime for ONNX (class OnnxInference), visualization tools (see Visualization), and a numpy API for ONNX). The package also provides tools to compare predictions, to benchmark models converted with sklearn-onnx.
import numpy
from sklearn.linear_model import LinearRegression
from sklearn.datasets import load_iris
from mlprodict.onnxrt import OnnxInference
from mlprodict.onnxrt.validate.validate_difference import measure_relative_difference
from mlprodict import __max_supported_opset__, get_ir_version
iris = load_iris()
X = iris.data[:, :2]
y = iris.target
lr = LinearRegression()
lr.fit(X, y)
# Predictions with scikit-learn.
expected = lr.predict(X[:5])
print(expected)
# Conversion into ONNX.
from mlprodict.onnx_conv import to_onnx
model_onnx = to_onnx(lr, X.astype(numpy.float32),
black_op={'LinearRegressor'},
target_opset=__max_supported_opset__)
print("ONNX:", str(model_onnx)[:200] + "\n...")
# Predictions with onnxruntime
model_onnx.ir_version = get_ir_version(__max_supported_opset__)
oinf = OnnxInference(model_onnx, runtime='onnxruntime1')
ypred = oinf.run({'X': X[:5].astype(numpy.float32)})
print("ONNX output:", ypred)
# Measuring the maximum difference.
print("max abs diff:", measure_relative_difference(expected, ypred['variable']))
# And the python runtime
oinf = OnnxInference(model_onnx, runtime='python')
ypred = oinf.run({'X': X[:5].astype(numpy.float32)},
verbose=1, fLOG=print)
print("ONNX output:", ypred)Installation
Installation from pip should work unless you need the latest development features.
pip install mlprodictThe package includes a runtime for ONNX. That's why there is a limited number of dependencies. However, some features relies on sklearn-onnx, onnxruntime, scikit-learn. They can be installed with the following instructions:
pip install mlprodict[all]The code is available at GitHub/mlprodict and has online documentation <http://www.xavierdupre.fr/app/ mlprodict/helpsphinx/index.html>.