A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 15, 22:15 UTC
[ROCM] fix native attention function call (#13650)
Gordon Wong committed
Feb 22, 2025 at 14:07 UTC
68d630a0c7273777c23a89e3abab75bd89eaa644
1 file changed
-1
vllm/attention/backends/rocm_flash_attn.py
-1
@@ -717,7 +717,6 @@ class ROCmFlashAttentionImpl(AttentionImpl):
717
self.num_heads,
718
self.head_size,
719
self.scale,
720
- causal_mask,
720
attn_masks,
721
)
722
else: