token bug

app.py

@@ -38,7 +38,7 @@ class WebSocketStreamingCallbackHandler(BaseCallbackHandler):
 from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
 
 class ChatHuggingFace:
-    def __init__(self, model, token, temperature=0.3, streaming=False):
+    def __init__(self, model, token="", temperature=0.3, streaming=False):
         # Instead of using InferenceClient, load the model locally.
         self.temperature = temperature
         self.streaming = streaming
@@ -104,7 +104,7 @@ model_name="Qwen/Qwen2.5-1.5B-Instruct"
 # llm = ChatOllama(model=model_name, temperature=0.3, streaming=True)
 llm = ChatHuggingFace(
     model=model_name,
-    # token=token,
+    # token=token
     temperature=0.3,
     streaming=True  # or True, based on your needs
 )