41 lines
1.2 KiB
Python
41 lines
1.2 KiB
Python
# Copyright (C) 2018-2021 Intel Corporation
|
|
# SPDX-License-Identifier: Apache-2.0
|
|
|
|
import numpy as np
|
|
import pytest
|
|
|
|
import ngraph as ng
|
|
from tests.test_ngraph.util import run_op_node
|
|
|
|
|
|
@pytest.mark.parametrize(
|
|
"shape_a, shape_b, transpose_a, transpose_b",
|
|
[
|
|
# matrix, vector
|
|
([2, 4], [4], False, False),
|
|
([4], [4, 2], False, False),
|
|
# matrix, matrix
|
|
([2, 4], [4, 2], False, False),
|
|
# tensor, vector
|
|
([2, 4, 5], [5], False, False),
|
|
# # tensor, matrix
|
|
([2, 4, 5], [5, 4], False, False),
|
|
# # tensor, tensor
|
|
([2, 2, 4], [2, 4, 2], False, False),
|
|
],
|
|
)
|
|
def test_matmul(shape_a, shape_b, transpose_a, transpose_b):
|
|
np.random.seed(133391)
|
|
left_input = -100.0 + np.random.rand(*shape_a).astype(np.float32) * 200.0
|
|
right_input = -100.0 + np.random.rand(*shape_b).astype(np.float32) * 200.0
|
|
|
|
result = run_op_node([left_input, right_input], ng.matmul, transpose_a, transpose_b)
|
|
|
|
if transpose_a:
|
|
left_input = np.transpose(left_input)
|
|
if transpose_b:
|
|
right_input = np.transpose(right_input)
|
|
|
|
expected = np.matmul(left_input, right_input)
|
|
assert np.allclose(result, expected)
|