onnx_light.onnx.helper

Functions

get_attribute_value(attr)	Returns the attribute value whatever the type is.
make_attribute(key, value[, doc_string, ...])	Makes an AttributeProto based on the value type.
make_attribute_ref(name, attr_type[, doc_string])	Makes an AttributeProto holding a reference to the parent " function's attribute of given name and type.
make_empty_tensor_value_info(name)	Creates an empty tensor value info.
make_function(domain, fname, inputs, ...[, ...])	Constructs a FunctionProto.
make_graph(nodes, name, inputs, outputs[, ...])	Constructs a GraphProto
make_map(name, key_type, keys, values)	Make a Map with specified key-value pair arguments.
make_map_type_proto(key_type, value_type)	Makes a map TypeProto.
make_model(graph[, ir_version, ...])	Constructs a ModelProto
make_node(op_type, inputs, outputs[, name, ...])	Constructs a NodeProto.
make_operatorsetid(domain, version)	Construct an OperatorSetIdProto.
make_opsetid(domain, version)	Constructs an OperatorSetIdProto.
make_optional(name, elem_type, value)	Make an Optional with specified value arguments.
make_optional_type_proto(inner_type_proto)	Makes an optional TypeProto.
make_sequence(name, elem_type, values)	Make a Sequence with specified value arguments.
make_sequence_type_proto(inner_type_proto)	Makes a sequence TypeProto.
make_sparse_tensor(values, indices, dims)	Construct a SparseTensorProto
make_sparse_tensor_type_proto(elem_type, shape)	Makes a SparseTensor TypeProto based on the data type and shape.
make_sparse_tensor_value_info(name, ...[, ...])	Makes a SparseTensor ValueInfoProto based on the data type and shape.
make_tensor(name, data_type, dims, vals[, raw])	Makes a TensorProto with specified arguments.
make_tensor_sequence_value_info(name, ...[, ...])	Makes a Sequence[Tensors] ValueInfoProto based on the data type and shape.
make_tensor_type_proto(elem_type, shape[, ...])	Makes a Tensor TypeProto based on the data type and shape.
make_tensor_value_info(name, elem_type, shape)	Makes a ValueInfoProto based on the data type and shape.
make_value_info(name, type_proto[, doc_string])	Makes a ValueInfoProto with the given type_proto.
set_metadata_props(proto, dict_value)	Sets metadata_props.
set_model_props(model, dict_value)	Sets metadata_props.
tensor_dtype_to_field(tensor_dtype)	Converts a TensorProto's data_type to corresponding field name for storage.
tensor_dtype_to_np_dtype(tensor_dtype)	Converts a TensorProto's data_type to corresponding numpy dtype.

Classes

TensorDtypeMap(np_dtype, storage_dtype, name)