fix streamer

Author: IlyasMoutawwakil

optimum/intel/openvino/modeling_decoder.py

@@ -56,6 +56,11 @@
 
 
 if TYPE_CHECKING:
+    try:
+        from transformers.generation.streamers import BaseStreamer
+    except Exception:
+        from typing import Generator as BaseStreamer
+
     from transformers.modeling_utils import PreTrainedModel
 
 
@@ -705,6 +710,7 @@ def generate(
         prefix_allowed_tokens_fn: Optional[Callable[[int, torch.Tensor], List[int]]] = None,
         synced_gpus: Optional[bool] = None,
         assistant_model: Optional["PreTrainedModel"] = None,
+        streamer: Optional["BaseStreamer"] = None,
         negative_prompt_ids: Optional[torch.Tensor] = None,
         negative_prompt_attention_mask: Optional[torch.Tensor] = None,
         **kwargs,
@@ -732,6 +738,7 @@ def generate(
             prefix_allowed_tokens_fn,
             synced_gpus,
             assistant_model,
+            streamer,
             negative_prompt_ids,
             negative_prompt_attention_mask,
             **kwargs,