Use is_flashinfer_available to replace is_hip for flashinfer check (#1596)
Co-authored-by: Zhang Liangang <liangang.zhang@intel.com>
This commit is contained in:
@@ -47,10 +47,9 @@ from sglang.srt.layers.quantization.base_config import QuantizationConfig
|
||||
from sglang.srt.layers.radix_attention import RadixAttention
|
||||
from sglang.srt.managers.schedule_batch import global_server_args_dict
|
||||
from sglang.srt.model_executor.forward_batch_info import ForwardBatch
|
||||
from sglang.srt.utils import is_hip
|
||||
from sglang.srt.utils import is_flashinfer_available
|
||||
|
||||
# ROCm: flashinfer available later
|
||||
if not is_hip():
|
||||
if is_flashinfer_available():
|
||||
from flashinfer import bmm_fp8
|
||||
|
||||
|
||||
|
||||
@@ -43,10 +43,9 @@ from sglang.srt.layers.quantization.base_config import QuantizationConfig
|
||||
from sglang.srt.layers.radix_attention import RadixAttention
|
||||
from sglang.srt.managers.schedule_batch import global_server_args_dict
|
||||
from sglang.srt.model_executor.forward_batch_info import ForwardBatch
|
||||
from sglang.srt.utils import is_hip
|
||||
from sglang.srt.utils import is_flashinfer_available
|
||||
|
||||
# ROCm: flashinfer available later
|
||||
if not is_hip():
|
||||
if is_flashinfer_available():
|
||||
from flashinfer import bmm_fp8
|
||||
|
||||
|
||||
|
||||
Reference in New Issue
Block a user