pt: add 4 tabulate_fusion op (deepmodeling#3877)

**OP**: tabulate_multi_device.cc -> tabulate_fusion_se_a & atten & r & t
**Compile**: add compile file
**Test**: source/tests, se_t has no test, and all other ops pass the
forward and backward tests.

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

- **New Features**
- Introduced support for tabulating fusion operations for deep learning
models using PyTorch tensors.
- Added functions for forward, backward, and gradient computations for
various fusion operations, handling different data types and device
types.
- Registered new operations with PyTorch libraries for seamless
integration.

- **Chores**
- Updated build configuration to include new source files for fusion
operations.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->

---------

Signed-off-by: Jinzhe Zeng <jinzhe.zeng@rutgers.edu>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Jinzhe Zeng <jinzhe.zeng@rutgers.edu>

source/op/pt/CMakeLists.txt

@@ -1,4 +1,4 @@
-file(GLOB OP_SRC print_summary.cc comm.cc)
+file(GLOB OP_SRC print_summary.cc comm.cc tabulate_multi_device.cc)
 
 add_library(deepmd_op_pt MODULE ${OP_SRC})
 # link: libdeepmd libtorch