Merge branch 'v0.11.0-dev-wm-0122' into 'v0.11.0-dev'

[fix]解决dp+graph+naive all2all dispatch时的维度不匹配问题

See merge request dcutoolkit/deeplearing/vllm!384

vllm/v1/worker/gpu_model_runner.py

@@ -3026,6 +3026,8 @@ class GPUModelRunner(LoRAModelRunnerMixin, KVConnectorModelRunnerMixin):
             num_tokens_across_dp = num_tokens_after_padding
             num_tokens_after_padding = int(num_tokens_after_padding[0].item())
 
+        num_tokens = num_tokens_after_padding
+
         attn_metadata: Optional[PerLayerAttnMetadata] = None
 
         # If force_attention is True, we always capture attention. Otherwise,