[V1][Spec Decode] Change Spec Decode Rejection Sampling API (#13729)

vllm/v1/sample/metadata.py

@@ -13,9 +13,6 @@ class SamplingMetadata:
     all_greedy: bool
     all_random: bool
 
-    # None when there are no speculated tokens.
-    spec_token_ids: Optional[List[List[int]]]
-
     top_p: Optional[torch.Tensor]
     top_k: Optional[torch.Tensor]
     min_p: Optional[torch.Tensor]