Update app.py

app.py

@@ -1,7 +1,8 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
+from gradio_client import Client
 
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
+# Initialize the client with the new model
+client = Client("nikravan/Bitnet-1B")
 
 def respond(
     message,
@@ -24,26 +25,32 @@ def respond(
     response = ""
 
     # Get the model's response
-    for response_chunk in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = response_chunk.choices[0].delta.content
-        response += token
-
-    # Post-process the response to ensure relevance
-    if "Sorry, I can only answer questions related to the Constitution of India." in response:
-        response = "Sorry, I can only answer questions related to the Constitution of India."
+    try:
+        result = client.predict(
+            api_name="/chat",
+            messages=messages,
+            max_tokens=max_tokens,
+            temperature=temperature,
+            top_p=top_p
+        )
+        response = result['text']  # Adjust based on the actual output format
+
+        # Post-process the response to ensure relevance
+        if not is_constitution_related(response):
+            response = "Sorry, I can only answer questions related to the Constitution of India."
+
+    except Exception as e:
+        response = f"Error: {str(e)}"
 
     yield response
 
+def is_constitution_related(response):
+    # Perform a simple check to see if the response seems related to the Constitution
+    return "constitution" in response.lower() or "article" in response.lower()
+
 demo = gr.ChatInterface(
     respond,
     additional_inputs=[
-        gr.Textbox(value="You are a knowledgeable assistant specializing in the Constitution of India. Answer only questions related to the Constitution. If you find that a question is not relevant, inform the user accordingly.", label="System message"),
         gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
         gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
         gr.Slider(