Update app.py

app.py

@@ -1,6 +1,5 @@
 import gradio as gr
 from huggingface_hub import InferenceClient
-from typing import List, Tuple
 
 """
 For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
@@ -9,9 +8,9 @@ client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
 
 
 def respond(
-    message: str,
-    history: List[Tuple[str, str]],
-    system_message: str,
+    message,
+    history: list[tuple[str, str]],
+    system_message,
     max_tokens=1024,
     temperature=0.7,
     top_p=0.95,
@@ -36,13 +35,12 @@ def respond(
             temperature=temperature,
             top_p=top_p,
         ):
-            if message.choices and "delta" in message.choices[0]:
-                token = message.choices[0].delta.get("content", "")
-                response += token
-                yield response
-    except Exception as e:
-        print(f"Error during API call: {e}")
-        yield "An error occurred while processing the request."
+            token = message.choices[0].delta.content
+
+            response += token
+            yield response
+    except json.JSONDecodeError as e:
+        yield f"Error: {e}"
 
 """
 For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
@@ -65,4 +63,4 @@ demo = gr.ChatInterface(
 
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()