[V0 Deprecation] Remove V0 sampling metadata (#25345)

Signed-off-by: Woosuk Kwon <woosuk@thinkingmachines.ai>
This commit is contained in:
Woosuk Kwon
2025-09-21 10:37:11 -07:00
committed by GitHub
parent c438b2951c
commit 1c3ffdbecc
141 changed files with 172 additions and 583 deletions
+1 -4
View File
@@ -39,7 +39,6 @@ from vllm.config import VllmConfig
from vllm.distributed import parallel_state
from vllm.distributed import utils as dist_utils
from vllm.logger import init_logger
from vllm.model_executor import SamplingMetadata
from vllm.model_executor.layers.activation import QuickGELU
from vllm.model_executor.layers.layernorm import RMSNorm
from vllm.model_executor.layers.linear import (ColumnParallelLinear,
@@ -1292,11 +1291,9 @@ class Ernie4_5_VLMoeForConditionalGeneration(nn.Module, SupportsMultiModal,
def compute_logits(
self,
hidden_states: torch.Tensor,
sampling_metadata: SamplingMetadata,
) -> Optional[torch.Tensor]:
"""compute logits"""
return self.language_model.compute_logits(hidden_states,
sampling_metadata)
return self.language_model.compute_logits(hidden_states)
def _vision_forward(
self,