forked from mindspore-Ecosystem/mindspore
77 lines
3.0 KiB
Python
77 lines
3.0 KiB
Python
# Copyright 2020 Huawei Technologies Co., Ltd
|
|
#
|
|
# 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 numpy as np
|
|
|
|
import mindspore as ms
|
|
from mindspore import Tensor
|
|
from mindspore.parallel._utils import _to_full_shapes, _to_full_tensor
|
|
|
|
|
|
def test_to_full_shapes():
|
|
device_num = 16
|
|
shapes = [[32, 128], [12], [24, 1, 12]]
|
|
full_shapes = _to_full_shapes(shapes, device_num)
|
|
assert full_shapes == [(512, 128), (192,), (384, 1, 12)]
|
|
|
|
|
|
def test_to_full_tensor_1():
|
|
elem = Tensor([[1, 2, 3], [4, 5, 6]], dtype=ms.float32)
|
|
device_num = 4
|
|
global_rank = 2
|
|
full_tensor = _to_full_tensor(elem, device_num, global_rank, scaling_sens=None)
|
|
|
|
expect = ([[0, 0, 0], [0, 0, 0], [0, 0, 0], [0, 0, 0], [1, 2, 3], [4, 5, 6], [0, 0, 0], [0, 0, 0]])
|
|
expect_tensor = Tensor(expect, dtype=ms.float32)
|
|
|
|
assert np.all(full_tensor[0].asnumpy() == expect_tensor.asnumpy())
|
|
|
|
|
|
def test_to_full_tensor_2():
|
|
elem0 = Tensor([[1, 2, 3], [4, 5, 6]], dtype=ms.float32)
|
|
elem1 = Tensor([[1], [4]], dtype=ms.int32)
|
|
elem = (elem0, elem1,)
|
|
device_num = 4
|
|
global_rank = 2
|
|
full_tensor = _to_full_tensor(elem, device_num, global_rank, scaling_sens=None)
|
|
|
|
expect0 = ([[0, 0, 0], [0, 0, 0], [0, 0, 0], [0, 0, 0], [1, 2, 3], [4, 5, 6], [0, 0, 0], [0, 0, 0]])
|
|
expect_tensor0 = Tensor(expect0, dtype=ms.float32)
|
|
expect1 = ([[0], [0], [0], [0], [1], [4], [0], [0]])
|
|
expect_tensor1 = Tensor(expect1, dtype=ms.int32)
|
|
expect_tensors = (expect_tensor0, expect_tensor1)
|
|
|
|
assert np.all(full_tensor[0].asnumpy() == expect_tensors[0].asnumpy())
|
|
assert np.all(full_tensor[1].asnumpy() == expect_tensors[1].asnumpy())
|
|
|
|
|
|
def test_to_full_tensor_sens_2():
|
|
elem0 = Tensor([[1, 2, 3], [4, 5, 6]], dtype=ms.float32)
|
|
elem1 = Tensor([[1], [4]], dtype=ms.int32)
|
|
elem = (elem0, elem1,)
|
|
device_num = 4
|
|
global_rank = 2
|
|
full_tensor = _to_full_tensor(elem, device_num, global_rank, scaling_sens=0.1)
|
|
|
|
expect0 = ([[0, 0, 0], [0, 0, 0], [0, 0, 0], [0, 0, 0], [1, 2, 3], [4, 5, 6], [0, 0, 0], [0, 0, 0]])
|
|
expect_tensor0 = Tensor(expect0, dtype=ms.float32)
|
|
expect1 = ([[0], [0], [0], [0], [1], [4], [0], [0]])
|
|
expect_tensor1 = Tensor(expect1, dtype=ms.int32)
|
|
expect_tensor_sens = Tensor(0.1, dtype=ms.float32)
|
|
expect_tensors = (expect_tensor0, expect_tensor1, expect_tensor_sens)
|
|
|
|
assert np.all(full_tensor[0].asnumpy() == expect_tensors[0].asnumpy())
|
|
assert np.all(full_tensor[1].asnumpy() == expect_tensors[1].asnumpy())
|
|
assert np.all(full_tensor[2].asnumpy() == expect_tensors[2].asnumpy())
|