kanghui0204 5e634dd1bd feat: Low Precision Allreduce for PCIe based GPU (#3851) ... 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>		2025-05-14 16:45:43 +08:00
..
images	feat: Low Precision Allreduce for PCIe based GPU (#3851)	2025-05-14 16:45:43 +08:00
disaggregated-service.md	fix cache buffer (#3942)	2025-05-07 09:49:44 +08:00
executor.md	feat: Allow individual gatherContext for each additional output (#3374)	2025-04-12 17:00:36 +08:00
expert-parallelism.md	Update TensorRT-LLM (#2094)	2024-08-07 16:44:43 +08:00
gpt-attention.md	Update TensorRT-LLM (#2755)	2025-02-11 03:01:00 +00:00
gpt-runtime.md	Feat: Variable-Beam-Width-Search (VBWS) part4 (#3979)	2025-05-12 22:32:29 +02:00
graph-rewriting.md	Update TensorRT-LLM (#1492)	2024-04-24 14:44:22 +08:00
kv-cache-reuse.md	Update TensorRT-LLM (#2873)	2025-03-11 21:13:42 +08:00
lora.md	doc: fix path after examples migration (#3814)	2025-04-24 02:36:45 +08:00
lowprecision-pcie-allreduce.md	feat: Low Precision Allreduce for PCIe based GPU (#3851)	2025-05-14 16:45:43 +08:00
speculative-decoding.md	Doc: update steps of using Draft-Target-Model (DTM) in the documents. (#3366)	2025-04-09 17:35:01 +08:00
weight-streaming.md	doc: fix path after examples migration (#3814)	2025-04-24 02:36:45 +08:00