vllm/attention at 326976291b541f0fd5bef34aa1ff4a84bf8fb37d - vllm - 丝路新云-代码仓

xinyun/vllm

mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-04-11 18:27:03 +08:00

History

Chengji Yao 2a84fb422f

[TPU] kv cache update kernel doesn't need to be padded slices to multiple of num_slices_per_block (#22394 )

Signed-off-by: Chengji Yao <chengjiyao@gmail.com>
Co-authored-by: Chengji Yao <chengjiyao@gmail.com>

2025-08-09 20:49:04 -07:00

..

[Attention] Support multiple attention metadata builders per kv_cache_spec + proper local attention no hybrid kv cache fix (#21588 )

2025-08-06 18:40:52 -07:00

[Attention] Support multiple attention metadata builders per kv_cache_spec + proper local attention no hybrid kv cache fix (#21588 )

2025-08-06 18:40:52 -07:00

[TPU] kv cache update kernel doesn't need to be padded slices to multiple of num_slices_per_block (#22394 )

2025-08-09 20:49:04 -07:00

[MISC] Add init files for python package (#20908 )

2025-07-15 12:16:33 +00:00

__init__.py

[Misc] Add SPDX-FileCopyrightText (#19100 )

2025-06-03 11:20:17 -07:00

layer.py

[Attention] Support multiple attention metadata builders per kv_cache_spec + proper local attention no hybrid kv cache fix (#21588 )

2025-08-06 18:40:52 -07:00

selector.py

[Attention] Support multiple attention metadata builders per kv_cache_spec + proper local attention no hybrid kv cache fix (#21588 )

2025-08-06 18:40:52 -07:00