[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
+3 -4
View File
@@ -399,11 +399,10 @@ class ArceeForCausalLM(nn.Module, SupportsLoRA, SupportsPP):
inputs_embeds=inputs_embeds)
return model_output
def compute_logits(self, hidden_states: torch.Tensor,
sampling_metadata) -> Optional[torch.Tensor]:
def compute_logits(self,
hidden_states: torch.Tensor) -> Optional[torch.Tensor]:
# Compute final logits from hidden states (last pipeline rank only)
logits = self.logits_processor(self.lm_head, hidden_states,
sampling_metadata)
logits = self.logits_processor(self.lm_head, hidden_states)
return logits
def get_input_embeddings(self, input_ids: torch.Tensor) -> torch.Tensor: