Implementation:Microsoft Onnxruntime Convert Sklearn

Metadata

Field	Value
Implementation Name	Convert_Sklearn
Repository	Microsoft_Onnxruntime
Source Repository	https://github.com/microsoft/onnxruntime
Type	Wrapper Doc
Wrapper Tool	skl2onnx
Language	Python
Domain	ML_Inference, Model_Conversion
Last Updated	2026-02-10
Workflow	Train_Convert_Predict
Pair	3 of 5

Overview

Wrapper documentation for the skl2onnx.convert_sklearn() function, which transforms a trained scikit-learn model into ONNX format for inference with ONNX Runtime.

API Signature

skl2onnx.convert_sklearn(model, initial_types=initial_type) -> onnx.ModelProto

Import

from skl2onnx import convert_sklearn

Code Reference

Reference	Location
Usage example	docs/python/examples/plot_train_convert_predict.py:L56-58

I/O Contract

Inputs

Parameter	Type	Required	Description
`model`	scikit-learn estimator	Yes	A trained (fitted) scikit-learn model or pipeline.
`initial_types`	`list[tuple[str, FloatTensorType]]`	Yes	Input schema definition mapping tensor names to typed tensor descriptors.

Outputs

Output	Type	Description
return value	`onnx.ModelProto`	The converted ONNX model. Call `.SerializeToString()` to serialize to bytes for saving to disk.

Usage Example

from skl2onnx import convert_sklearn
from skl2onnx.common.data_types import FloatTensorType

# Define input schema
initial_type = [("float_input", FloatTensorType([None, 4]))]

# Convert trained model to ONNX
onx = convert_sklearn(clr, initial_types=initial_type)

# Save to disk
with open("model.onnx", "wb") as f:
    f.write(onx.SerializeToString())

From the source at docs/python/examples/plot_train_convert_predict.py:L56-58:

initial_type = [("float_input", FloatTensorType([None, 4]))]
onx = convert_sklearn(clr, initial_types=initial_type)
with open("logreg_iris.onnx", "wb") as f:
    f.write(onx.SerializeToString())

RandomForest Conversion

From docs/python/examples/plot_train_convert_predict.py:L183-186:

initial_type = [("float_input", FloatTensorType([1, 4]))]
onx = convert_sklearn(rf, initial_types=initial_type)
with open("rf_iris.onnx", "wb") as f:
    f.write(onx.SerializeToString())

Related Pages

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment