p60975123
/
mindspore
forked from mindspore-Ecosystem/mindspore
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

!10173 Fix Strategy Check for UnsortedSegments 

From: @huangxinjing
Reviewed-by: @zhunaipan,@stsuteng
Signed-off-by: @stsuteng
This commit is contained in:
mindspore-ci-bot 2020-12-19 14:24:58 +08:00 committed by Gitee
parent 615c74e596 a8446af1ab
commit 5ff6de83eb
2 changed files with 13 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
mindspore/ccsrc/frontend/parallel/ops_info/unsorted_segment_op_info.cc
View File

@ -87,7 +87,7 @@ Status UnsortedSegmentOpInfo::CheckStrategy(const StrategyPtr &strategy) {
Shape input_b_shape = inputs_shape_.at(1);
// The size of the input b must be equal or smaller than input a
for (size_t i = 0; i < input_b_shape.size(); ++i) {
if ((sub_a_strategy[i] != sub_b_strategy[i]) && (input_a_shape[i] != input_b_shape[i])) {
if ((sub_a_strategy[i] != sub_b_strategy[i]) || (input_a_shape[i] != input_b_shape[i])) {
MS_LOG(ERROR) << name_
<< " : Invalid strategy. The shape and the strategy of the input0 and input1 "
"should be same before the front size of the input[1]";

12
tests/ut/python/parallel/test_unsortedsegmentsum.py
View File

@ -14,6 +14,7 @@
# ============================================================================
import numpy as np
import pytest
import mindspore as ms
import mindspore.nn as nn
@ -112,6 +113,17 @@ def test_unsortedsegmentsum_model_parallel_index_slice_3d():
compile_graph(x, y, num_segments, strategy1, strategy2)
def test_unsortedsegmentsum_model_parallel_index_slice_diff_inputs():
context.set_auto_parallel_context(device_num=4, global_rank=0)
x = Tensor(np.ones((4, 4, 8)), ms.float32)
y = Tensor(np.ones((4, 4)), ms.int32)
num_segments = 16
strategy1 = (2, 2, 1)
strategy2 = (2, 4)
with pytest.raises(RuntimeError):
compile_graph(x, y, num_segments, strategy1, strategy2)
def test_unsortedsegmentsum_model_parallel_vector_slice_2d():
context.set_auto_parallel_context(device_num=4, global_rank=0)
x = Tensor(np.ones((4, 8)), ms.float32)