Files
openvino/model-optimizer/extensions/front/binary_quantize_normalization_test.py

76 lines
2.4 KiB
Python

"""
Copyright (C) 2018-2020 Intel Corporation
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
"""
import unittest
import numpy as np
from extensions.front.binary_quantize_normalization import BinaryFakeQuantizeNormalization
from mo.utils.ir_engine.compare_graphs import compare_graphs
from mo.utils.unittest.graph import build_graph
graph_nodes = {
'0': {'name': 'input', 'kind': 'op', 'op': 'Parameter'},
'1': {'name': 'mi_i', 'kind': 'op', 'op': 'Const'},
'2': {'name': 'ma_i', 'kind': 'op', 'op': 'Const'},
'3': {'name': 'mi_o', 'kind': 'op', 'op': 'Const'},
'4': {'name': 'mi_o', 'kind': 'op', 'op': 'Const'},
'add': {'kind': 'op', 'op': 'Add'},
'const': {'kind': 'op', 'op': 'Const', 'value': np.array(0.5)},
'mul': {'kind': 'op', 'op': 'Mul'},
'quantize': {'name': 'quantize', 'levels': 2, 'kind': 'op', 'op': 'FakeQuantize'},
'output': {'name': 'output1', 'kind': 'op', 'op': 'Result', 'type': 'Result'},
}
graph_edges = [
('0', 'quantize', {'in': 0}),
('1', 'quantize', {'in': 1}),
('2', 'quantize', {'in': 2}),
('3', 'quantize', {'in': 3}),
('4', 'quantize', {'in': 4}),
('quantize', 'output'),
]
graph_ref_edges = [
('0', 'quantize', {'in': 0}),
('1', 'add'),
('2', 'add'),
('add', 'mul'),
('const', 'mul'),
('mul', 'quantize', {'in': 1, 'out': 0}),
('mul', 'quantize', {'in': 2, 'out': 0}),
('3', 'quantize', {'in': 3}),
('4', 'quantize', {'in': 4}),
('quantize', 'output'),
]
class TestBinaryQuantizeNormalization(unittest.TestCase):
def test_binary_quantize_normalizer(self):
graph = build_graph(graph_nodes, graph_edges, nodes_with_edges_only=True)
graph.stage = 'front'
BinaryFakeQuantizeNormalization().find_and_replace_pattern(graph)
graph.clean_up()
graph_ref = build_graph(graph_nodes, graph_ref_edges)
graph_ref.clean_up()
(flag, resp) = compare_graphs(graph, graph_ref, 'output')
self.assertTrue(flag, resp)