mindspore/tests/st/ops/gpu/test_primitive_cache.py

119 lines
3.5 KiB
Python

# Copyright 2022 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.
# ============================================================================
""" test primitive cache """
import pytest
import numpy as np
import mindspore as ms
import mindspore.nn as nn
from mindspore import Tensor
from mindspore import context
from mindspore import jit
from mindspore.ops import operations as P
from mindspore.ops._primitive_cache import _get_cache_prim
# pylint: disable=W0235
@pytest.mark.level1
@pytest.mark.platform_x86_gpu_training
@pytest.mark.env_onecard
def test_jit_function_run_in_pynative():
"""
Feature: test @jit decorated function run in PyNative.
Description: test @jit decorated function run in PyNative.
Expectation: Success.
"""
context.set_context(mode=context.PYNATIVE_MODE, device_target='GPU')
@jit
def pow_function(x, y):
_pow = _get_cache_prim(P.Pow)()
return _pow(x, y)
class Pow(nn.Cell):
def __init__(self):
super(Pow, self).__init__()
def construct(self, x1, x2):
return pow_function(x1, x2)
x = Tensor(np.array([1.0, 2.0, 4.0]), ms.float32)
y = 3
output = Pow()(x, y)
expect_output = np.array([1.0, 8.0, 64.0], dtype=np.float32)
np.testing.assert_almost_equal(output.asnumpy(), expect_output)
@pytest.mark.level1
@pytest.mark.platform_x86_gpu_training
@pytest.mark.env_onecard
def test_run_pynative_and_then_run_graph():
"""
Feature: test the cache key must be a str.
Description: test run_pynative and then run_graph.
Expectation: Success.
"""
class Pow(nn.Cell):
def __init__(self):
super(Pow, self).__init__()
def construct(self, x1, x2):
_pow = _get_cache_prim(P.Pow)()
return _pow(x1, x2)
context.set_context(mode=context.PYNATIVE_MODE, device_target='GPU')
x = Tensor(np.array([1.0, 2.0, 4.0]), ms.float32)
y = 3
output1 = Pow()(x, y)
context.set_context(mode=context.GRAPH_MODE, device_target='GPU')
x = Tensor(np.array([1.0, 2.0, 4.0]), ms.float32)
y = 3
output2 = Pow()(x, y)
np.testing.assert_almost_equal(output1.asnumpy(), output2.asnumpy())
@pytest.mark.level1
@pytest.mark.platform_x86_gpu_training
@pytest.mark.env_onecard
def test_continuous_cache():
"""
Feature: test continuous cache.
Description: test continuous cache.
Expectation: Success.
"""
context.set_context(mode=context.PYNATIVE_MODE, device_target='GPU')
class AddSub(nn.Cell):
def __init__(self):
super(AddSub, self).__init__()
def construct(self, x, y):
y = y + 1
add = _get_cache_prim(P.Add)()
sub = _get_cache_prim(P.Sub)()
z = add(x, y)
out = sub(z, x)
return out
x = Tensor(np.array([2, 2, 1]), dtype=ms.int32)
y = Tensor(np.array([1, 1, 1]), dtype=ms.int32)
output = AddSub()(x, y)
expect_output = np.array([2, 2, 2], dtype=np.int32)
np.testing.assert_almost_equal(output.asnumpy(), expect_output)