| .. |
|
cubin
|
Feat: add sliding-window-attention generation-phase kernels on Blackwell (#4564)
|
2025-05-26 09:06:33 +08:00 |
|
decoderXQAImplJIT
|
feat : add PositionEmbeddingType=0 to xqa support (#4934)
|
2025-06-05 21:50:42 +08:00 |
|
instantiation
|
Update TensorRT-LLM (#2502)
|
2024-11-26 16:51:34 +08:00 |
|
CMakeLists.txt
|
infra: open source XQA kernels (#3762)
|
2025-04-30 18:05:15 +08:00 |
|
copy_cu.py
|
Update TensorRT-LLM (#787)
|
2024-01-02 17:54:32 +08:00 |
|
decoderMaskedMultiheadAttentionLaunch.h
|
[https://nvbugspro.nvidia.com/bug/5300080] Fix the bug of setting attention_chunk_size and enable chunked-attention in the generation-phase by default (#4693)
|
2025-06-03 19:02:57 -04:00 |
|
decoderMaskedMultiheadAttentionTemplate.h
|
[https://nvbugspro.nvidia.com/bug/5300080] Fix the bug of setting attention_chunk_size and enable chunked-attention in the generation-phase by default (#4693)
|
2025-06-03 19:02:57 -04:00 |
|
decoderXQAConstants.h
|
Update TensorRT-LLM (#2094)
|
2024-08-07 16:44:43 +08:00 |
|
decoderXQAImpl.cpp
|
Update TensorRT-LLM (#2783)
|
2025-02-13 18:40:22 +08:00 |
|
decoderXQAImpl.h
|
chore: remove usernames from comments (#3291)
|
2025-04-05 13:44:28 +08:00 |
|
decoderXQAImplCommon.cpp
|
Support speculative decoding with Hopper XQA (#3269)
|
2025-04-07 17:14:34 +08:00 |
|
decoderXQAImplCommon.h
|
Support speculative decoding with Hopper XQA (#3269)
|
2025-04-07 17:14:34 +08:00 |
|
decoderXQAImplPrecompiled.cpp
|
feat: add CGA reduction fmha kernels on Blackwell. (#3763)
|
2025-04-29 10:43:54 +08:00 |
|
decoderXQAImplPrecompiled.h
|
Update TensorRT-LLM (#2783)
|
2025-02-13 18:40:22 +08:00 |
|
decoderXQARunner.cpp
|
Support speculative decoding with Hopper XQA (#3269)
|
2025-04-07 17:14:34 +08:00 |
|
decoderXQARunner.h
|
Update TensorRT-LLM (#2873)
|
2025-03-11 21:13:42 +08:00 |
|
mmha_notes.md
|
Initial commit
|
2023-09-20 00:29:41 -07:00 |
|
tensorMapUtils.cpp
|
Update TensorRT-LLM (#2783)
|
2025-02-13 18:40:22 +08:00 |
|
tensorMapUtils.h
|
Update TensorRT-LLM (#1688)
|
2024-05-28 20:07:49 +08:00 |
|
xqaParams.h
|
[https://nvbugspro.nvidia.com/bug/5300080] Fix the bug of setting attention_chunk_size and enable chunked-attention in the generation-phase by default (#4693)
|
2025-06-03 19:02:57 -04:00 |
