diff --git a/docs/design/attention_backends.md b/docs/design/attention_backends.md index 3d0fcd6c77..6d5c007e3e 100644 --- a/docs/design/attention_backends.md +++ b/docs/design/attention_backends.md @@ -172,7 +172,7 @@ Priority is **1 = highest** (tried first). | `FLEX_ATTENTION` | | fp16, bf16, fp32 | `auto`, `bfloat16` | Any | Any | ❌ | ✅ | ❌ | Decoder, Encoder Only | Any | | `ROCM_AITER_FA` | | fp16, bf16 | `auto` | 16, 32 | 64, 128, 256 | ❌ | ❌ | ❌ | Decoder | N/A | | `ROCM_AITER_UNIFIED_ATTN` | | fp16, bf16 | `auto` | Any | Any | ❌ | ❌ | ❌ | All | N/A | -| `ROCM_ATTN` | | fp16, bf16, fp32 | `auto` | 16, 32, 544 | 32, 64, 96, 128, 160, 192, 224, 256 | ❌ | ❌ | ❌ | All | N/A | +| `ROCM_ATTN` | | fp16, bf16, fp32 | `auto` | 16, 32, 544 | 32, 64, 80, 96, 128, 160, 192, 224, 256 | ❌ | ❌ | ❌ | All | N/A | | `TREE_ATTN` | | fp16, bf16 | `auto` | %16 | 32, 64, 96, 128, 160, 192, 224, 256 | ❌ | ❌ | ❌ | Decoder | Any | | `TRITON_ATTN` | | fp16, bf16, fp32 | `auto`, `bfloat16`, `fp8`, `fp8_e4m3`, `fp8_e5m2` | %16 | Any | ✅ | ✅ | ❌ | All | Any | diff --git a/vllm/v1/attention/backends/rocm_attn.py b/vllm/v1/attention/backends/rocm_attn.py index d4bfa764fe..b53170c989 100644 --- a/vllm/v1/attention/backends/rocm_attn.py +++ b/vllm/v1/attention/backends/rocm_attn.py @@ -182,7 +182,7 @@ class RocmAttentionBackend(AttentionBackend): @classmethod def get_supported_head_sizes(cls) -> list[int]: - return [32, 64, 96, 128, 160, 192, 224, 256] + return [32, 64, 80, 96, 128, 160, 192, 224, 256] @classmethod def validate_head_size(cls, head_size: int) -> None: