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

python/cutlass/emit/pytorch.py

@@ -39,7 +39,7 @@ Example usage with JIT compilation:
 .. highlight:: python
 .. code-block:: python
 
-    plan = cutlass.op.Gemm(element=torch.float32, layout=cutlass.LayoutType.RowMajor)
+    plan = cutlass.op.Gemm(element=torch.float32, layout=cutlass_library.LayoutType.RowMajor)
     op = plan.construct()
     mod = cutlass.emit.pytorch(op, 'cutlass_gemm', 80, jit=True)
 
@@ -81,15 +81,16 @@ The module can later be used in Python via:
 import logging
 import os
 
-from cutlass import CUTLASS_PATH, logger, swizzle, ConvKind, ConvKindNames, DataType
+from cutlass_library import ConvKind, ConvKindNames, DataType, SubstituteTemplate
+
+from cutlass import CUTLASS_PATH, logger, swizzle
 from cutlass.backend.gemm_operation import GemmOperationGrouped, GemmOperationUniversal
 from cutlass.backend.conv2d_operation import Conv2dOperation
 from cutlass.backend.library import ApiVersion
-from cutlass.backend.utils.software import CheckPackages, SubstituteTemplate
 from cutlass.emit import common
+from cutlass.utils.datatypes import is_torch_available
 
-torch_available = CheckPackages().check_torch()
-if torch_available:
+if is_torch_available():
     import torch