louis_lifu
/
vllm
mirror of https://github.com/vllm-project/vllm
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Fix: MI100 Support By Bypassing Custom Paged Attention (#9560)

This commit is contained in:
ErkinSagiroglu 2024-10-26 13:12:57 +01:00 committed by GitHub
parent 5cbdccd151
commit 55137e8ee3
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 6 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
vllm/attention/backends/rocm_flash_attn.py
View File

@ -21,7 +21,10 @@ if TYPE_CHECKING:
logger = init_logger(__name__)
_PARTITION_SIZE_ROCM = 512
_ON_NAVI = "gfx1" in torch.cuda.get_device_properties("cuda").gcnArchName
_GPU_ARCH = torch.cuda.get_device_properties("cuda").gcnArchName
_ON_NAVI = "gfx1" in _GPU_ARCH
_ON_MI250_MI300 = any(arch in _GPU_ARCH
for arch in ["gfx90a", "gfx940", "gfx941", "gfx942"])
class ROCmFlashAttentionBackend(AttentionBackend):
@ -662,7 +665,8 @@ def _use_rocm_custom_paged_attention(qtype: torch.dtype, head_size: int,
block_size: int, gqa_ratio: int,
max_seq_len: int) -> bool:
# rocm custom page attention not support on navi (gfx1*)
return (not _ON_NAVI and (qtype == torch.half or qtype == torch.bfloat16)
return (_ON_MI250_MI300 and not _ON_NAVI
and (qtype == torch.half or qtype == torch.bfloat16)
and (head_size == 64 or head_size == 128)
and (block_size == 16 or block_size == 32)
and (gqa_ratio >= 1 and gqa_ratio <= 16) and max_seq_len <= 32768)