fix

Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>

Author: LucasWilkinson

vllm/v1/worker/gpu_model_runner.py

@@ -707,7 +707,7 @@ def _prepare_inputs(
 
             # Fill unused with -1. Needed for reshape_and_cache in full cuda
             # graph mode.
-            slot_mapping[total_num_scheduled_tokens:].fill_(-1)
+            blk_table.slot_mapping[total_num_scheduled_tokens:].fill_(-1)
 
             common_attn_metadata = CommonAttentionMetadata(
                 query_start_loc=self.query_start_loc[:num_reqs + 1],