Skip to content

P
Paddle

BaiXuePrincess / Paddle
与 Fork 源项目一致

Fork自 PaddlePaddle / Paddle

通知 1
Star 1
Fork 0
P
Paddle
未验证 提交 3cab25a5 编写于 作者: X Xin Pan 提交者： GitHub
浏览文件
操作

Merge pull request #14062 from tensor-tang/fix/jit/avx 

fix avx error
上级 d0fdcb2f a05fce65
显示空白变更内容
内联 并排
Showing with 2 addition and 0 deletion
paddle/fluid/operators/math/jit_kernel_rnn.cc
浏览文件 @ 3cab25a5
...@@ -136,6 +136,7 @@ static std::shared_ptr<const VActKernel<T>> GetActKernel( ...@@ -136,6 +136,7 @@ static std::shared_ptr<const VActKernel<T>> GetActKernel(
return nullptr; return nullptr;
} }
#ifdef __AVX__
template <jit::cpu_isa_t isa> template <jit::cpu_isa_t isa>
static std::unique_ptr<AVXAct> GetAVXAct(const std::string& type) { static std::unique_ptr<AVXAct> GetAVXAct(const std::string& type) {
if (type == "sigmoid") { if (type == "sigmoid") {
...@@ -150,6 +151,7 @@ static std::unique_ptr<AVXAct> GetAVXAct(const std::string& type) { ...@@ -150,6 +151,7 @@ static std::unique_ptr<AVXAct> GetAVXAct(const std::string& type) {
PADDLE_THROW("Not support type: %s", type); PADDLE_THROW("Not support type: %s", type);
return nullptr; return nullptr;
} }
#endif
/* LSTM JitKernel */ /* LSTM JitKernel */
template <typename T, jit::cpu_isa_t isa, jit_block> template <typename T, jit::cpu_isa_t isa, jit_block>
......
Markdown is supported
0% 或 .
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 或