CUTLASS 3.3.0 (#1167)

* Release 3.3.0

Adds support for mixed precision GEMMs On Hopper and Ampere
Adds support for < 16B aligned GEMMs on Hopper
Enhancements to EVT
Enhancements to Python interface
Enhancements to Sub-byte type handling in CuTe
Several other bug-fixes and performance improvements.

* minor doc update

tools/library/include/cutlass/library/operation_table.h

@@ -215,21 +215,21 @@ struct GemmPreferenceKey {
     return compute_capability == rhs.compute_capability;
   }
 };
+
 /////////////////////////////////////////////////////////////////////////////////////////////////
+
 inline
 std::ostream& operator<< (std::ostream& out, const cutlass::library::GemmPreferenceKey& key) {
     out << "{\n"
       << "compute_capability : " << key.compute_capability << std::endl
       << "alignment          : " << key.alignment << std::endl
       << "}";
-  
+
   return out;
 }
 
 /////////////////////////////////////////////////////////////////////////////////////////////////
 
-/////////////////////////////////////////////////////////////////////////////////////////////////
-
 /// Maps minimum compute capability onto a vector of possible operations
 using GemmOperationVectorMap = std::map<
   GemmPreferenceKey,
@@ -242,7 +242,6 @@ using GemmOperationFunctionalMap = std::unordered_map<
   GemmOperationVectorMap,
   GemmFunctionalKeyHasher
 >;
-/////////////////////////////////////////////////////////////////////////////////////////////////
 
 /////////////////////////////////////////////////////////////////////////////////////////////////
 //                          Data Structures for Conv Functional Maps