diff --git a/modeling/inference_model.py b/modeling/inference_model.py
index 1af06675..441335cc 100644
--- a/modeling/inference_model.py
+++ b/modeling/inference_model.py
@@ -46,8 +46,8 @@ def __enter__(self):
             )
 
         if use_core_manipulations.sample:
-            use_core_manipulations.old_sample = transformers.GenerationMixin.sample
-            transformers.GenerationMixin.sample = use_core_manipulations.sample
+            use_core_manipulations.old_sample = transformers.GenerationMixin._sample
+            transformers.GenerationMixin._sample = use_core_manipulations.sample
 
         if use_core_manipulations.get_stopping_criteria:
             use_core_manipulations.old_get_stopping_criteria = (
@@ -69,7 +69,7 @@ def __exit__(self, exc_type, exc_value, exc_traceback):
             ), "Patch leak: THE MONKEYS HAVE ESCAPED"
 
         if use_core_manipulations.old_sample:
-            transformers.GenerationMixin.sample = use_core_manipulations.old_sample
+            transformers.GenerationMixin._sample = use_core_manipulations.old_sample
         else:
             assert (
                 not use_core_manipulations.sample
diff --git a/modeling/inference_models/hf_torch.py b/modeling/inference_models/hf_torch.py
index fcdd9fb9..37eaf105 100644
--- a/modeling/inference_models/hf_torch.py
+++ b/modeling/inference_models/hf_torch.py
@@ -266,7 +266,7 @@ def new_sample(self, *args, **kwargs):
                 kwargs.setdefault("pad_token_id", 2)
             return new_sample.old_sample(self, *args, **kwargs)
 
-        new_sample.old_sample = transformers.GenerationMixin.sample
+        new_sample.old_sample = transformers.GenerationMixin._sample
         use_core_manipulations.sample = new_sample
 
         # PEFT Loading. This MUST be done after all save_pretrained calls are