[Bugfix][Kernel] Fix CUDA 11.8 being broken by FA3 build (#12375)
Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
This commit is contained in:
Regular → Executable
+11
-3
@@ -18,17 +18,20 @@ from vllm.attention.backends.utils import (
|
||||
get_seq_len_block_table_args, is_all_cross_attn_metadata_set,
|
||||
is_all_encoder_attn_metadata_set, is_block_tables_empty)
|
||||
from vllm.envs import VLLM_FLASH_ATTN_VERSION
|
||||
from vllm.logger import init_logger
|
||||
from vllm.multimodal import MultiModalPlaceholderMap
|
||||
from vllm.platforms import current_platform
|
||||
from vllm.utils import async_tensor_h2d, make_tensor_with_pad
|
||||
from vllm.vllm_flash_attn import (fa_version_unsupported_reason,
|
||||
flash_attn_varlen_func,
|
||||
flash_attn_with_kvcache,
|
||||
is_fa_version_supported)
|
||||
|
||||
if TYPE_CHECKING:
|
||||
from vllm.worker.model_runner import (ModelInputForGPUBuilder,
|
||||
ModelInputForGPUWithSamplingMetadata)
|
||||
|
||||
from vllm.vllm_flash_attn import (flash_attn_varlen_func,
|
||||
flash_attn_with_kvcache,
|
||||
is_fa_version_supported)
|
||||
logger = init_logger(__name__)
|
||||
|
||||
|
||||
class FlashAttentionBackend(AttentionBackend):
|
||||
@@ -652,6 +655,11 @@ class FlashAttentionImpl(AttentionImpl):
|
||||
assert VLLM_FLASH_ATTN_VERSION in [2, 3]
|
||||
self.fa_version = VLLM_FLASH_ATTN_VERSION
|
||||
|
||||
if not is_fa_version_supported(self.fa_version):
|
||||
logger.error("Cannot use FA version %d is not supported due to %s",
|
||||
self.fa_version,
|
||||
fa_version_unsupported_reason(self.fa_version))
|
||||
|
||||
assert is_fa_version_supported(self.fa_version)
|
||||
|
||||
def forward(
|
||||
|
||||
Regular → Executable
+10
-1
@@ -10,11 +10,15 @@ import triton.language as tl
|
||||
from vllm.attention.backends.abstract import (AttentionBackend, AttentionImpl,
|
||||
AttentionMetadata, AttentionType)
|
||||
from vllm.envs import VLLM_FLASH_ATTN_VERSION
|
||||
from vllm.logger import init_logger
|
||||
from vllm.platforms import current_platform
|
||||
from vllm.utils import cdiv
|
||||
from vllm.vllm_flash_attn import (flash_attn_varlen_func,
|
||||
from vllm.vllm_flash_attn import (fa_version_unsupported_reason,
|
||||
flash_attn_varlen_func,
|
||||
is_fa_version_supported)
|
||||
|
||||
logger = init_logger(__name__)
|
||||
|
||||
|
||||
class FlashAttentionBackend(AttentionBackend):
|
||||
|
||||
@@ -143,6 +147,11 @@ class FlashAttentionImpl(AttentionImpl):
|
||||
assert VLLM_FLASH_ATTN_VERSION in [2, 3]
|
||||
self.fa_version = VLLM_FLASH_ATTN_VERSION
|
||||
|
||||
if not is_fa_version_supported(self.fa_version):
|
||||
logger.error("Cannot use FA version %d is not supported due to %s",
|
||||
self.fa_version,
|
||||
fa_version_unsupported_reason(self.fa_version))
|
||||
|
||||
assert is_fa_version_supported(self.fa_version)
|
||||
|
||||
def forward(
|
||||
|
||||
Reference in New Issue
Block a user