mindspore-Ecosystem
/
mindspore
21
0
Fork 27
Code Issues Pull Requests 6 Projects Releases Wiki Activity

!14967 fix weight quant conv bug 

From: @xutianchun
Reviewed-by: @HilbertDavid,@ivss
Signed-off-by: @HilbertDavid
This commit is contained in:
mindspore-ci-bot 2021-04-13 10:22:41 +08:00 committed by Gitee
parent 16fa94f45a 84959e8baf
commit 1d501a85c4
1 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
mindspore/lite/tools/converter/quantizer/weight_quantizer.cc
View File

@ -100,8 +100,8 @@ STATUS WeightQuantizer::DoConvQuantize(const CNodePtr &cnode) {
}
if (tensor_info->data_type() != mindspore::kNumberTypeFloat32) {
MS_LOG(ERROR) << "model weight data type invalid which is " << tensor_info->data_type();
return RET_ERROR;
MS_LOG(WARNING) << cnode->fullname_with_scope() << " weight data type is not fp32 but " << tensor_info->data_type();
return RET_OK;
}
auto status = RET_ERROR;
if (type_id_ == kNumberTypeInt8) {