"vllm/v1/attention/backend.py" did not exist on "8fcaaf6a165e661f63fc51be906bc05b0767332f"
- 23 Apr, 2026 3 commits
- 22 Apr, 2026 2 commits
- 18 Apr, 2026 2 commits
- 17 Apr, 2026 1 commit
-
-
王敏 authored
-
- 16 Apr, 2026 1 commit
-
-
chenhw5 authored
-
- 15 Apr, 2026 1 commit
-
-
wanghl6 authored
-
- 11 Apr, 2026 1 commit
-
-
laibao authored
-
- 10 Apr, 2026 3 commits
- 08 Apr, 2026 2 commits
- 03 Apr, 2026 2 commits
- 02 Apr, 2026 2 commits
- 01 Apr, 2026 3 commits
- 28 Mar, 2026 1 commit
-
-
wanglong3 authored
-
- 27 Mar, 2026 3 commits
-
-
flyingdown authored
-
laibao authored
-
flyingdown authored
-
- 26 Mar, 2026 6 commits
-
-
laibao authored
-
laibao authored
feat(v1 attention): 为 ROCm FlashAttention 接入 unified kv layout，并打通 mm_prefix、qq_bias 与 use_alibi_sqrt 透传 在 ROCm FlashAttention 后端增加 unified KV layout 选择逻辑 接入 unified varlen kernel 调用路径 在 FlashAttention metadata 中补充 mm_prefix_range 与 qq_bias 透传
-
wanghl6 authored
-
wanghl6 authored
-
wanghl6 authored
-
wanglong3 authored
-
- 24 Mar, 2026 6 commits
- 23 Mar, 2026 1 commit
-
-
guanyu1 authored
-
