[Bugfix] Remove num_tokens_across_dp (#14302)

Signed-off-by: Tyler Michael Smith <[email protected]>

Author: tlrmchlsmth

vllm/forward_context.py

@@ -27,7 +27,6 @@
 
 @dataclass
 class DPMetadata:
-    num_tokens_across_dp: list[int]
     cu_tokens_across_dp_cpu: torch.Tensor
 
 
@@ -89,7 +88,7 @@ def set_forward_context(attn_metadata: Any,
         from vllm.distributed.parallel_state import get_dp_group
         dist.all_reduce(num_tokens_tensor, group=get_dp_group().cpu_group)
         cu_tokens_across_dp_cpu = torch.cumsum(num_tokens_tensor, dim=0)
-        dp_metadata = DPMetadata(num_tokens_across_dp, cu_tokens_across_dp_cpu)
+        dp_metadata = DPMetadata(cu_tokens_across_dp_cpu)
 
     global _forward_context
     prev_context = _forward_context