abetlen
diff --git a/‎llama_cpp/llama.py
Copy file name to clipboardExpand all lines: llama_cpp/llama.py
+6Lines changed: 6 additions & 0 deletions b/‎llama_cpp/llama.py
Copy file name to clipboardExpand all lines: llama_cpp/llama.py
+6Lines changed: 6 additions & 0 deletions
@@ -1034,6 +1034,7 @@ def _create_completion(
         logits_processor: Optional[LogitsProcessorList] = None,
         grammar: Optional[LlamaGrammar] = None,
         logit_bias: Optional[Dict[str, float]] = None,
+        reset: bool = True,
     ) -> Union[
         Iterator[CreateCompletionResponse], Iterator[CreateCompletionStreamResponse]
     ]:
@@ -1222,6 +1223,7 @@ def logit_bias_processor(
             stopping_criteria=stopping_criteria,
             logits_processor=logits_processor,
             grammar=grammar,
+            reset=reset,
         ):
             assert self._model.model is not None
             if llama_cpp.llama_token_is_eog(self._model.model, token):
@@ -1653,6 +1655,7 @@ def create_completion(
         logits_processor: Optional[LogitsProcessorList] = None,
         grammar: Optional[LlamaGrammar] = None,
         logit_bias: Optional[Dict[str, float]] = None,
+        reset: bool = True,
     ) -> Union[CreateCompletionResponse, Iterator[CreateCompletionStreamResponse]]:
         """Generate text from a prompt.
 
@@ -1716,6 +1719,7 @@ def create_completion(
             logits_processor=logits_processor,
             grammar=grammar,
             logit_bias=logit_bias,
+            reset=reset,
         )
         if stream:
             chunks: Iterator[CreateCompletionStreamResponse] = completion_or_chunks
@@ -1746,6 +1750,7 @@ def __call__(
         mirostat_tau: float = 5.0,
         mirostat_eta: float = 0.1,
         model: Optional[str] = None,
+        reset: bool = True,
         stopping_criteria: Optional[StoppingCriteriaList] = None,
         logits_processor: Optional[LogitsProcessorList] = None,
         grammar: Optional[LlamaGrammar] = None,
@@ -1813,6 +1818,7 @@ def __call__(
             logits_processor=logits_processor,
             grammar=grammar,
             logit_bias=logit_bias,
+            reset=reset,
         )
 
     def create_chat_completion(