fix: add types

load_models.py

@@ -211,9 +211,6 @@ def load_model(device_type, model_id, model_basename=None, LOGGING=logging, stre
 
     # Create a pipeline for text generation
 
-
-    streamer = TextStreamer(tokenizer, skip_prompt=True)
-
     pipe = pipeline(
         "text-generation",
         model=model,
@@ -224,11 +221,10 @@ def load_model(device_type, model_id, model_basename=None, LOGGING=logging, stre
         top_k=40,
         repetition_penalty=1.0,
         generation_config=generation_config,
-        streamer=streamer,
         # callbacks=callbacks
     )
 
     local_llm = HuggingFacePipeline(pipeline=pipe)
     logging.info("Local LLM Loaded")
 
-    return [local_llm, streamer]
+    return local_llm