kanghui0204 6f3922f318 feat: Low Precision Allreduce for PCIe based GPU (#4344) ... This PR adds a customized allreduce to TensorRT-LLM. The new allreduce is used for communication on PCIe-based GPUs via low-precision quantization, which can accelerate the PCIe allreduce process. Signed-off-by: Hui Kang <hkang@nvidia.com> Co-authored-by: Hui Kang <hkang@nvidia.com>		2025-05-20 06:53:46 +08:00
..
__init__.py	Update TensorRT-LLM (#1019)	2024-01-31 21:55:32 +08:00
plugin.py	feat: Low Precision Allreduce for PCIe based GPU (#4344)	2025-05-20 06:53:46 +08:00