onnx.shape_inference

infer_shapes

onnx.shape_inference.infer_shapes(model: ModelProto | bytes, check_type: bool = False, strict_mode: bool = False, data_prop: bool = False) → ModelProto

Apply shape inference to the provided ModelProto.

Inferred shapes are added to the value_info field of the graph.

If the inferred values conflict with values already provided in the graph, that means that the provided values are invalid (or there is a bug in shape inference), and the result is unspecified.

Parameters:

• model – ModelProto.

• check_type – Checks the type-equality for input and output.

• strict_mode – Stricter shape inference, it will throw errors if any; Otherwise, simply stop if any error.

• data_prop – Enables data propagation for limited operators to perform shape computation.

Returns:

(ModelProto) model with inferred shape information

infer_shapes_path

onnx.shape_inference.infer_shapes_path(model_path: str | PathLike, output_path: str | PathLike = '', check_type: bool = False, strict_mode: bool = False, data_prop: bool = False) → None

Take model path for shape_inference.

This function is the same as infer_shape() but supports >2GB models. The function outputs the inferred model to the output_path. The original model path is used if not specified.

infer_node_outputs

onnx.shape_inference.infer_node_outputs(schema: OpSchema, node: NodeProto, input_types: dict[str, TypeProto], input_data: dict[str, TensorProto] | None = None, input_sparse_data: dict[str, SparseTensorProto] | None = None, opset_imports: list[OperatorSetIdProto] | None = None, ir_version: int = 11) → dict[str, TypeProto]

infer_function_output_types

onnx.shape_inference.infer_function_output_types(function: FunctionProto, input_types: Sequence[TypeProto], attributes: Sequence[AttributeProto]) → list[TypeProto]

Apply type-and-shape-inference to given function body, with given input types and given input attribute values.