wehub-resource-sync
101 lines
3.5 KiB
Python
101 lines
3.5 KiB
Python
# Copyright (c) 2023 PaddlePaddle Authors. All Rights Reserved.
|
|
#
|
|
# 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
|
|
from semi_auto_parallel_util import SemiAutoParallelTestBase
|
|
|
|
import paddle
|
|
import paddle.distributed as dist
|
|
from paddle.distributed import Replicate, Shard
|
|
|
|
"""
|
|
test for reshape
|
|
"""
|
|
|
|
|
|
class TestReshapeSemiAutoParallel(SemiAutoParallelTestBase):
|
|
def __init__(self):
|
|
super().__init__()
|
|
|
|
def check_placements(self, output, expected_placements):
|
|
assert output.placements == expected_placements, (
|
|
f"{output.placements} vs {expected_placements}"
|
|
)
|
|
|
|
def test_reshape_forward(self):
|
|
shape = [200, 30]
|
|
mesh = dist.ProcessMesh([0, 1], dim_names=["x"])
|
|
input = dist.shard_tensor(
|
|
paddle.rand(shape=[10, 20, 30]),
|
|
mesh,
|
|
[Shard(0), Replicate(), Replicate()],
|
|
)
|
|
input.stop_gradient = False
|
|
output = paddle.reshape(input, shape)
|
|
output.backward()
|
|
|
|
self.check_placements(output, [dist.Shard(0)])
|
|
self.check_placements(input.grad, [dist.Shard(0)])
|
|
|
|
def test_reshape_infer_shape(self):
|
|
mesh = dist.ProcessMesh([0, 1], dim_names=["x"])
|
|
x = paddle.ones([10, 20, 30])
|
|
x = dist.shard_tensor(x, mesh, [Shard(0)])
|
|
y = x.reshape([-1, 0, x.shape[0]])
|
|
assert y.shape == [30, 20, 10]
|
|
assert y._local_shape == [15, 20, 10]
|
|
|
|
def test_shape_api_with_reshape(self):
|
|
mesh = dist.ProcessMesh([0, 1], dim_names=["x"])
|
|
a = paddle.rand(shape=[4, 6, 8])
|
|
b = dist.shard_tensor(a, mesh, [dist.Shard(0)])
|
|
|
|
dist_shape = paddle.shape(b)
|
|
b = b.reshape((-1, dist_shape[-1]))
|
|
assert b.shape == [24, 8]
|
|
assert b._local_shape == [12, 8]
|
|
|
|
def test_reshape_grad_with_reshard_x_grad(self):
|
|
mesh = dist.ProcessMesh([0, 1], dim_names=["x"])
|
|
|
|
x = paddle.rand(shape=[2, 8, 4, 48])
|
|
x.stop_gradient = False
|
|
dist_x = dist.shard_tensor(x, mesh, [Shard(2)])
|
|
dist_out = dist_x.reshape([64, 48]) # dist_x needs reshard
|
|
|
|
# calling reshape_grad, its output dist_x.grad is replicated
|
|
# on the mesh, different from dist_x's placements,
|
|
# so it will reshard to dist_x's placements [Shard(2)]
|
|
dist_out.backward()
|
|
|
|
np.testing.assert_equal(dist_x._local_shape, [2, 8, 2, 48])
|
|
np.testing.assert_equal(dist_out._local_shape, [64, 48])
|
|
np.testing.assert_equal(dist_x.grad._local_shape, [2, 8, 2, 48])
|
|
|
|
def run_test_case(self):
|
|
if self._backend == "cpu":
|
|
paddle.set_device("cpu")
|
|
elif self._backend == "gpu":
|
|
paddle.set_device("gpu:" + str(dist.get_rank()))
|
|
else:
|
|
raise ValueError("Only support cpu or gpu backend.")
|
|
self.test_reshape_forward()
|
|
self.test_reshape_infer_shape()
|
|
self.test_shape_api_with_reshape()
|
|
self.test_reshape_grad_with_reshard_x_grad()
|
|
|
|
|
|
if __name__ == '__main__':
|
|
TestReshapeSemiAutoParallel().run_test_case()
|