Fix triton compilation issue (#3984)

Co-authored-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
2024-04-12 16:41:26 -07:00
parent fbb9d9eef4
commit d04973ad54
1 changed files with 5 additions and 1 deletions
--- a/vllm/attention/ops/triton_flash_attention.py
+++ b/vllm/attention/ops/triton_flash_attention.py
@@ -415,7 +415,11 @@ def attn_fwd(
            return

    is_mqa = hq != hk
-    off_h_k = off_h_q % hk if is_mqa else off_h_q
+    if is_mqa:  # noqa: SIM108
+        off_h_k = off_h_q % hk
+    else:
+        off_h_k = off_h_q
+
    n_extra_tokens = 0
    if seqlen_k < BLOCK_N:
        n_extra_tokens = BLOCK_N - seqlen_k