diff --git a/src/transformers/generation/utils.py b/src/transformers/generation/utils.py
index ebe968b7ac4..51019da9a6b 100644
--- a/src/transformers/generation/utils.py
+++ b/src/transformers/generation/utils.py
@@ -1542,10 +1542,7 @@ class GenerationMixin:
             logger.warning(f"Setting `pad_token_id` to `eos_token_id`:{pad_token_tensor} for open-end generation.")
 
         # we can't infer attn mask if pad token is set to be eos token in model's generation config
-        if (
-            eos_token_tensor is not None
-            and torch.isin(elements=eos_token_tensor, test_elements=pad_token_tensor).any()
-        ):
+        if eos_token_tensor is not None and pad_token_tensor in eos_token_tensor:
             if kwargs_has_attention_mask is not None and not kwargs_has_attention_mask:
                 logger.warning_once(
                     "The attention mask is not set and cannot be inferred from input because pad token is same as eos token."