Binyang Li 28a57b0610 NVLS support for msccl++ executor (#375) ...

- Support mote datatype for multicast operation
- Add new OP MULTI_LOAD_REDUCE_STORE to support NVLS
- Modify allocSharedPhysicalCuda, which return std::shared_ptr<T>
instead of std::shared_ptr<PhysicalCudaMemory>
- Add Python support for allocSharedPhysicalCuda

Test passed for `allreduce_nvls.json`

2024-11-20 06:43:28 +00:00

..

__init__.py

NVLS support for msccl++ executor (#375)

2024-11-20 06:43:28 +00:00

CMakeLists.txt

Fix bootstrapping mechanism (#278)

2024-03-27 10:24:24 +08:00

comm.py

Auto-tune vector sizes for NVLS allreduce6 (#338)

2024-08-16 11:11:54 +08:00

core_py.cpp

NVLS support for msccl++ executor (#375)

2024-11-20 06:43:28 +00:00

error_py.cpp

Add executor to execute schedule-plan file (#283)

2024-04-18 19:10:41 +00:00

executor_py.cpp

Tune threads per block for mscclpp executor (#345)

2024-09-18 17:21:47 -07:00

fifo_py.cpp

Update interface to let user change fifo size (#243)

2024-01-09 22:14:36 -08:00

gpu_utils_py.cpp

NVLS support for msccl++ executor (#375)

2024-11-20 06:43:28 +00:00

npkit_py.cpp

Add NPKit GPU event support (#310)

2024-06-13 13:59:50 +08:00

numa_py.cpp

Pytest (#162)

2023-09-01 21:22:11 +08:00

nvls_py.cpp

NVLS support for msccl++ executor (#375)

2024-11-20 06:43:28 +00:00

proxy_channel_py.cpp

Update interface to let user change fifo size (#243)

2024-01-09 22:14:36 -08:00

semaphore_py.cpp

Release GIL for Python APIs with wait (#190)

2023-11-14 21:11:01 +08:00

sm_channel_py.cpp

Pytest (#162)

2023-09-01 21:22:11 +08:00

utils_py.cpp

NVLS support. (#250)

2024-02-04 20:46:10 -08:00

utils.py

Allow obtaining cuda stream handle from PyTorch stream when launching kernel (#297)

2024-05-04 04:57:07 +00:00