diff --git a/tensorrt_llm/_torch/speculative/dspark.py b/tensorrt_llm/_torch/speculative/dspark.py
index f2597f5e9ef3..80425833b0be 100644
--- a/tensorrt_llm/_torch/speculative/dspark.py
+++ b/tensorrt_llm/_torch/speculative/dspark.py
@@ -404,7 +404,7 @@ def _draft_gen_block_batched(
         )
         return block_logits
 
-    def forward(
+    def _forward_impl(
         self,
         input_ids,
         position_ids,
diff --git a/tests/unittest/_torch/speculative/test_rejection_buffers_guard.py b/tests/unittest/_torch/speculative/test_rejection_buffers_guard.py
index e2cbe8752d7a..a71b9f1e6640 100644
--- a/tests/unittest/_torch/speculative/test_rejection_buffers_guard.py
+++ b/tests/unittest/_torch/speculative/test_rejection_buffers_guard.py
@@ -201,6 +201,9 @@ class _Worker(SpecWorkerBase):
     def max_draft_len(self) -> int:
         return K
 
+    def _forward_impl(self, *args: object, **kwargs: object) -> None:
+        raise NotImplementedError
+
 
 def _dispatch_meta(**over):
     base = dict(