ONNX 模型分析与使用

import onnx# onnx_model is an in-mempry ModelProtoonnx_model = onnx.load('path/to/the/model.onnx') # 加载 onnx 模型

import onnxfrom onnx.external_data_helper import load_external_data_for_model
onnx_model = onnx.load('path/to/the/model.onnx', load_external_data=False)load_external_data_for_model(onnx_model, 'data/directory/path/')# Then the onnx_model has loaded the external data from the specific directory

from onnx.external_data_helper import convert_model_to_external_data
# onnx_model is an in-memory ModelProtoonnx_model = ...convert_model_to_external_data(onnx_model, all_tensors_to_one_file=True, location='filename', size_threshold=1024, convert_attribute=False)# Then the onnx_model has converted raw data as external data# Must be followed by save

import onnx
# onnx_model is an in-memory ModelProtoonnx_model = ...
# Save the ONNX modelonnx.save(onnx_model, 'path/to/the/model.onnx')

import onnx
# onnx_model is an in-memory ModelProtoonnx_model = ...onnx.save_model(onnx_model, 'path/to/save/the/model.onnx', save_as_external_data=True, all_tensors_to_one_file=True, location='filename', size_threshold=1024, convert_attribute=False)# Then the onnx_model has converted raw data as external data and saved to specific directory

import numpyimport onnxfrom onnx import numpy_helper
# Preprocessing: create a Numpy arraynumpy_array = numpy.array([[1.0, 2.0, 3.0], [4.0, 5.0, 6.0]], dtype=float)print('Original Numpy array:\n{}\n'.format(numpy_array))
# Convert the Numpy array to a TensorPrototensor = numpy_helper.from_array(numpy_array)print('TensorProto:\n{}'.format(tensor))
# Convert the TensorProto to a Numpy arraynew_array = numpy_helper.to_array(tensor)print('After round trip, Numpy array:\n{}\n'.format(new_array))
# Save the TensorProtowith open('tensor.pb', 'wb') as f:    f.write(tensor.SerializeToString())
# Load a TensorProtonew_tensor = onnx.TensorProto()with open('tensor.pb', 'rb') as f:    new_tensor.ParseFromString(f.read())print('After saving and loading, new TensorProto:\n{}'.format(new_tensor))

import onnxfrom onnx import helperfrom onnx import AttributeProto, TensorProto, GraphProto# The protobuf definition can be found here:# https://github.com/onnx/onnx/blob/master/onnx/onnx.proto
# Create one input (ValueInfoProto)X = helper.make_tensor_value_info('X', TensorProto.FLOAT, [3, 2])pads = helper.make_tensor_value_info('pads', TensorProto.FLOAT, [1, 4])
value = helper.make_tensor_value_info('value', AttributeProto.FLOAT, [1])
# Create one output (ValueInfoProto)Y = helper.make_tensor_value_info('Y', TensorProto.FLOAT, [3, 4])
# Create a node (NodeProto) - This is based on Pad-11node_def = helper.make_node(    'Pad',                  # name    ['X', 'pads', 'value'], # inputs    ['Y'],                  # outputs    mode='constant',        # attributes)
# Create the graph (GraphProto)graph_def = helper.make_graph(    [node_def],        # nodes    'test-model',      # name    [X, pads, value],  # inputs    [Y],               # outputs)
# Create the model (ModelProto)model_def = helper.make_model(graph_def, producer_name='onnx-example')
print('The model is:\n{}'.format(model_def))onnx.checker.check_model(model_def)print('The model is checked!')

import onnx
# Preprocessing: load the ONNX modelmodel_path = 'path/to/the/model.onnx'onnx_model = onnx.load(model_path)
print('The model is:\n{}'.format(onnx_model))
# Check the modeltry:    onnx.checker.check_model(onnx_model)except onnx.checker.ValidationError as e:    print('The model is invalid: %s' % e)else:    print('The model is valid!')

import onnx
input_path = 'path/to/the/original/model.onnx'output_path = 'path/to/save/the/extracted/model.onnx'input_names = ['input_0', 'input_1', 'input_2']output_names = ['output_0', 'output_1']
onnx.utils.extract_model(input_path, output_path, input_names, output_names)

import onnxfrom onnx.tools import update_model_dims
model = onnx.load('path/to/the/model.onnx')# Here both 'seq', 'batch' and -1 are dynamic using dim_param.variable_length_model = update_model_dims.update_inputs_outputs_dims(model, {'input_name': ['seq', 'batch', 3, -1]}, {'output_name': ['seq', 'batch', 1, -1]})# need to check model after the input/output sizes are updatedonnx.checker.check_model(variable_length_model )