[https://nvbugs/5698581][fix] Init draft tokens for CUDA graph dummy request (NVIDIA#9505)

Signed-off-by: ziyixiong-nv <219238287+ziyixiong-nv@users.noreply.github.com>

Author: ziyixiong-nv

tensorrt_llm/_torch/pyexecutor/cuda_graph_runner.py

@@ -389,6 +389,7 @@ def _get_padded_batch(self, batch: ScheduledRequests,
             if spec_res_mgr:
                 spec_res_mgr.add_dummy_requests([CUDA_GRAPH_DUMMY_REQUEST_ID])
 
+        self.padding_dummy_request.py_draft_tokens = [0] * runtime_draft_len
         batch.generation_requests.extend([self.padding_dummy_request] *
                                          padding_size)
         return padding_size