[Bugfix] Register VLLM_BATCH_INVARIANT in envs.py to fix spurious unknown env var warning (#35007)
Signed-off-by: Ranran <1012869439@qq.com> Signed-off-by: Ranran <hzz5361@psu.edu> Signed-off-by: ran <hzz5361@psu.edu> Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>
This commit is contained in:
@@ -7,6 +7,7 @@ from abc import abstractmethod
|
||||
import torch
|
||||
from torch.nn.parameter import Parameter, UninitializedParameter
|
||||
|
||||
import vllm.envs as envs
|
||||
from vllm.distributed import (
|
||||
divide,
|
||||
get_tensor_model_parallel_rank,
|
||||
@@ -19,7 +20,6 @@ from vllm.logger import init_logger
|
||||
from vllm.model_executor.custom_op import PluggableLayer
|
||||
from vllm.model_executor.layers.batch_invariant import (
|
||||
linear_batch_invariant,
|
||||
vllm_is_batch_invariant,
|
||||
)
|
||||
from vllm.model_executor.layers.quantization.base_config import (
|
||||
QuantizationConfig,
|
||||
@@ -223,7 +223,7 @@ class UnquantizedLinearMethod(LinearMethodBase):
|
||||
x: torch.Tensor,
|
||||
bias: torch.Tensor | None = None,
|
||||
) -> torch.Tensor:
|
||||
if vllm_is_batch_invariant() and current_platform.is_cuda_alike():
|
||||
if envs.VLLM_BATCH_INVARIANT and current_platform.is_cuda_alike():
|
||||
return linear_batch_invariant(x, layer.weight, bias)
|
||||
return dispatch_unquantized_gemm()(layer, x, layer.weight, bias)
|
||||
|
||||
|
||||
Reference in New Issue
Block a user