Merge with ZHANG Yuhao branch (Don't use this! It's an incomplete version)

Minor error fixed.

source/tensor/core/arithmetic/Sum.cpp

@@ -90,9 +90,8 @@ void _Sum(const XTensor * a, const XTensor * b, XTensor * c, DTYPE beta)
                 /* when c != a, OpenBLAS needs to copy a to c first. This operation
                 /* when c != a, OpenBLAS needs to copy a to c first. This operation
                  slow down the speed, so just use OpenBLAS when c == a */
                  slow down the speed, so just use OpenBLAS when c == a */
 #if defined(USE_BLAS)
 #if defined(USE_BLAS)
-                if (c == a) {
+                if (c == a) 
                     AXPY(a->unitNum,beta,bp,1,cp,1);
                     AXPY(a->unitNum,beta,bp,1,cp,1);
-                }
                 else {
                 else {
                     int num = a->unitNum;
                     int num = a->unitNum;
                     if (num % 4 == 0) {
                     if (num % 4 == 0) {