kanshan/TensorRT-LLMs
1
0
Fork 0
mirror of https://github.com/NVIDIA/TensorRT-LLM.git synced 2026-01-14 06:27:45 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity
2967d299fb
TensorRT-LLMs/cpp/tensorrt_llm/kernels/weightOnlyBatchedGemv
History
Cheng Hang 656c705ff1
[None][feat] sm100 weight-only kernel (#10190) 
Signed-off-by: Cheng Hang <chang@nvidia.com>
2026-01-05 09:44:36 +08:00
..
common.h [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
converter.h [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
cudaCoreGemm.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
cudaCoreGemm.h [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
cudaCoreGemmNVFP4.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
cudaCoreGemmNVFP4.h [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
details.h [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
int8SQ.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
int8SQ.h [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernel.h [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernelDispatcher.h [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernelDispatcherBf16Int4GroupwiseColumnMajorFalse.cu [None][feat] sm100 weight-only kernel (#10190) 2026-01-05 09:44:36 +08:00
kernelDispatcherBf16Int4GroupwiseColumnMajorInterleavedForHopperTrue.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernelDispatcherBf16Int4GroupwiseColumnMajorInterleavedTrue.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernelDispatcherBf16Int4PerChannelColumnMajorFalse.cu [None][feat] sm100 weight-only kernel (#10190) 2026-01-05 09:44:36 +08:00
kernelDispatcherBf16Int4PerChannelColumnMajorInterleavedForHopperTrue.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernelDispatcherBf16Int4PerChannelColumnMajorInterleavedTrue.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernelDispatcherBf16Int8GroupwiseColumnMajoInterleavedForHopperTrue.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernelDispatcherBf16Int8GroupwiseColumnMajorFalse.cu [None][feat] sm100 weight-only kernel (#10190) 2026-01-05 09:44:36 +08:00
kernelDispatcherBf16Int8GroupwiseColumnMajorInterleavedTrue.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernelDispatcherBf16Int8PerChannelColumnMajorFalse.cu [None][feat] sm100 weight-only kernel (#10190) 2026-01-05 09:44:36 +08:00
kernelDispatcherBf16Int8PerChannelColumnMajorInterleavedForHopperTrue.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernelDispatcherBf16Int8PerChannelColumnMajorInterleavedTrue.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernelDispatcherFp16Int4GroupwiseColumnMajorFalse.cu [None][feat] sm100 weight-only kernel (#10190) 2026-01-05 09:44:36 +08:00
kernelDispatcherFp16Int4GroupwiseColumnMajorInterleavedForHopperTrue.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernelDispatcherFp16Int4GroupwiseColumnMajorInterleavedTrue.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernelDispatcherFp16Int4PerChannelColumnMajorFalse.cu [None][feat] sm100 weight-only kernel (#10190) 2026-01-05 09:44:36 +08:00
kernelDispatcherFp16Int4PerChannelColumnMajorInterleavedForHopperTrue.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernelDispatcherFp16Int4PerChannelColumnMajorInterleavedTrue.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernelDispatcherFp16Int8GroupwiseColumnMajorFalse.cu [None][feat] sm100 weight-only kernel (#10190) 2026-01-05 09:44:36 +08:00
kernelDispatcherFp16Int8GroupwiseColumnMajorInterleavedForHopperTrue.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernelDispatcherFp16Int8GroupwiseColumnMajorInterleavedTrue.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernelDispatcherFp16Int8PerChannelColumnMajorFalse.cu [None][feat] sm100 weight-only kernel (#10190) 2026-01-05 09:44:36 +08:00
kernelDispatcherFp16Int8PerChannelColumnMajorInterleavedForHopperTrue.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernelDispatcherFp16Int8PerChannelColumnMajorInterleavedTrue.cu [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00
kernelLauncher.h [None][feat] sm100 weight-only kernel (#10190) 2026-01-05 09:44:36 +08:00
utility.h [None][fix] Introduce inline namespace to avoid symbol collision (#9541) 2025-12-12 23:32:15 +08:00