未验证 提交 42549fb7 编写于 作者: W whs 提交者: GitHub

Merge pull request #7811 from wanghaoshuang/fix_im2seq

Fix LoD calculate of im2seq
......@@ -79,7 +79,7 @@ class Im2SequenceKernel : public framework::OpKernel<T> {
framework::LoD lod(1);
lod[0].reserve(batch_size + 1);
for (int i = 0, offset = 0; i < batch_size + 1; ++i) {
lod[0][i] = offset;
lod[0].push_back(offset);
offset += output_height * output_width;
}
out->set_lod(lod);
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册