quark.onnx.optimizations.convert_transforms

Module Contents

Classes

class quark.onnx.optimizations.convert_transforms.ConvQDQToQOPTransform

Defines a transform to be applied to a onnx model graph.

A transform is a combination of ‘Find + Replace’ which describes how to find a pattern of nodes in a model, and what to replace those nodes with.

A pattern is described using OpTypePattern. The replacement function receives a NodeTree which contains the matched nodes and should return a NodeTree which contains the set of nodes which replaced the matched nodes.

pattern() → OpTypePattern

Return the OpTypePattern to find in the model graph.

replacement(match_node: NodeTree) → Any

Generate a replacement sub-graph for the matched sub-graph.

The fundamental constraint of the replacement is that the replacement sub-graph should consume the same input tensors as the original sub-graph and also produce a final list of tensors which are same in number and shape as the original sub-graph. Not following this could crash model creation, or introduce bugs in the new model graph.

Parameters:

match_nodes – Matched NodeTree based on self.pattern().

class quark.onnx.optimizations.convert_transforms.MatMulQDQToQOPTransform

Defines a transform to be applied to a onnx model graph.

A transform is a combination of ‘Find + Replace’ which describes how to find a pattern of nodes in a model, and what to replace those nodes with.

A pattern is described using OpTypePattern. The replacement function receives a NodeTree which contains the matched nodes and should return a NodeTree which contains the set of nodes which replaced the matched nodes.

pattern() → OpTypePattern

Return the OpTypePattern to find in the model graph.

replacement(match_node: NodeTree) → Any

Generate a replacement sub-graph for the matched sub-graph.

The fundamental constraint of the replacement is that the replacement sub-graph should consume the same input tensors as the original sub-graph and also produce a final list of tensors which are same in number and shape as the original sub-graph. Not following this could crash model creation, or introduce bugs in the new model graph.

Parameters:

match_nodes – Matched NodeTree based on self.pattern().

class quark.onnx.optimizations.convert_transforms.AddQDQToQOPTransform

Defines a transform to be applied to a onnx model graph.

A transform is a combination of ‘Find + Replace’ which describes how to find a pattern of nodes in a model, and what to replace those nodes with.

A pattern is described using OpTypePattern. The replacement function receives a NodeTree which contains the matched nodes and should return a NodeTree which contains the set of nodes which replaced the matched nodes.

pattern() → OpTypePattern

Return the OpTypePattern to find in the model graph.

replacement(match_node: NodeTree) → Any

Generate a replacement sub-graph for the matched sub-graph.

The fundamental constraint of the replacement is that the replacement sub-graph should consume the same input tensors as the original sub-graph and also produce a final list of tensors which are same in number and shape as the original sub-graph. Not following this could crash model creation, or introduce bugs in the new model graph.

Parameters:

match_nodes – Matched NodeTree based on self.pattern().

class quark.onnx.optimizations.convert_transforms.MulQDQToQOPTransform

Defines a transform to be applied to a onnx model graph.

A transform is a combination of ‘Find + Replace’ which describes how to find a pattern of nodes in a model, and what to replace those nodes with.

A pattern is described using OpTypePattern. The replacement function receives a NodeTree which contains the matched nodes and should return a NodeTree which contains the set of nodes which replaced the matched nodes.

pattern() → OpTypePattern

Return the OpTypePattern to find in the model graph.

replacement(match_node: NodeTree) → Any

Generate a replacement sub-graph for the matched sub-graph.

The fundamental constraint of the replacement is that the replacement sub-graph should consume the same input tensors as the original sub-graph and also produce a final list of tensors which are same in number and shape as the original sub-graph. Not following this could crash model creation, or introduce bugs in the new model graph.

Parameters:

match_nodes – Matched NodeTree based on self.pattern().

class quark.onnx.optimizations.convert_transforms.SigmoidQDQToQOPTransform

Defines a transform to be applied to a onnx model graph.

A transform is a combination of ‘Find + Replace’ which describes how to find a pattern of nodes in a model, and what to replace those nodes with.

A pattern is described using OpTypePattern. The replacement function receives a NodeTree which contains the matched nodes and should return a NodeTree which contains the set of nodes which replaced the matched nodes.

pattern() → OpTypePattern

Return the OpTypePattern to find in the model graph.

replacement(match_node: NodeTree) → Any

Generate a replacement sub-graph for the matched sub-graph.

The fundamental constraint of the replacement is that the replacement sub-graph should consume the same input tensors as the original sub-graph and also produce a final list of tensors which are same in number and shape as the original sub-graph. Not following this could crash model creation, or introduce bugs in the new model graph.

Parameters:

match_nodes – Matched NodeTree based on self.pattern().

class quark.onnx.optimizations.convert_transforms.RemoveQDQTransform

Defines a transform to be applied to a onnx model graph.

A transform is a combination of ‘Find + Replace’ which describes how to find a pattern of nodes in a model, and what to replace those nodes with.

A pattern is described using OpTypePattern. The replacement function receives a NodeTree which contains the matched nodes and should return a NodeTree which contains the set of nodes which replaced the matched nodes.

pattern() → OpTypePattern

Return the OpTypePattern to find in the model graph.

replacement(match_node: NodeTree) → Any

Generate a replacement sub-graph for the matched sub-graph.

The fundamental constraint of the replacement is that the replacement sub-graph should consume the same input tensors as the original sub-graph and also produce a final list of tensors which are same in number and shape as the original sub-graph. Not following this could crash model creation, or introduce bugs in the new model graph.

Parameters:

match_nodes – Matched NodeTree based on self.pattern().