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

fix incorrect in_channel size error messages

This commit is contained in:
tom__chen 2021-02-05 09:53:51 -05:00
parent 09db51a797
commit d3330a1087
2 changed files with 6 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
mindspore/core/abstract/prim_nn.cc
View File

@ -333,9 +333,10 @@ AbstractBasePtr InferImplConv2D(const AnalysisEnginePtr &, const PrimitivePtr &p
w_axis = 2;
}
int64_t group = CheckAttrPositiveInt64(op_name, primitive->GetAttr("group"), "group");
if ((x_shape[c_axis] != Shape::SHP_ANY) && (x_shape[c_axis] % group != 0)) {
MS_LOG(EXCEPTION) << "x_shape[" << c_axis << "] = " << x_shape[c_axis]
<< " (channels) must be divisible by group = " << group;
if ((x_shape[c_axis] != Shape::SHP_ANY) && (w_shape[c_axis] != Shape::SHP_ANY) &&
((x_shape[c_axis] / group) != w_shape[c_axis])) {
MS_LOG(EXCEPTION) << "x_shape[C_in] / group must equal to w_shape[C_in] = " << w_shape[c_axis] << ", but got "
<< (x_shape[c_axis] / group);
}
int64_t out_channel = CheckAttrPositiveInt64(op_name, primitive->GetAttr("out_channel"), "out_channel");
if ((w_shape[n_axis] != Shape::SHP_ANY) && (w_shape[n_axis] != out_channel)) {

3
tests/ut/python/model/test_lenet_core_after_exception.py
View File

@ -53,5 +53,6 @@ def test_lenet5_exception():
predict = Tensor(in1)
label = Tensor(in2)
net = train_step_with_loss_warp(LeNet5())
with pytest.raises(ValueError):
with pytest.raises(RuntimeError) as info:
_executor.compile(net, predict, label)
assert "x_shape[C_in] / group must equal to w_shape[C_in] = " in str(info.value)