kanshan/TensorRT-LLMs
1
0
Fork 0
mirror of https://github.com/NVIDIA/TensorRT-LLM.git synced 2026-02-04 18:21:52 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

[None][chore] Revert NVIDIA/TensorRT-LLM#10847 (#10869)

Browse Source
This commit is contained in:
Yanchao Lu 2026-01-21 11:08:40 +08:00 committed by GitHub
parent c381790d15
commit ccf4d79c6c
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
tensorrt_llm/_torch/modules/attention.py
View File

@ -334,11 +334,11 @@ class Attention(nn.Module):
key="sparse_attention_config")
if config.sparse_attention_config.algorithm == "rocket":
logger.info_once("disable rope_fusion for RocketKV.")
logger.warning("disable rope_fusion for RocketKV.")
self.rope_fusion = False
if self.rope_fusion and not attn_cls.support_fused_rope():
logger.info_once(
logger.warning(
"rope_fusion is true but the attention backend does not support it. Will disable rope_fusion."
)
self.rope_fusion = False