module onnxrt.optim.onnx_optimisation

Short summary

module mlprodict.onnxrt.optim.onnx_optimisation

Optimisations of ONNX graphs.

source on GitHub

Functions

function	truncated documentation
onnx_remove_node	Removes as many nodes as possible without changing the outcome. It applies onnx_remove_node_identity(), then …

Documentation

Optimisations of ONNX graphs.

source on GitHub

mlprodict.onnxrt.optim.onnx_optimisation.onnx_remove_node(onnx_model, recursive=True, debug_info=None, **options)

Removes as many nodes as possible without changing the outcome. It applies onnx_remove_node_identity, then onnx_remove_node_redundant.

Parameters

• onnx_model – onnx model

• recursive – looks into subgraphs

• debug_info – debug information (private)

• options – additional options

Returns

new onnx _model

source on GitHub