Update app.py

app.py

@@ -7,56 +7,31 @@ from mistralai.models.chat_completion import ChatMessage
 #client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
 client = MistralClient(api_key= os.getenv('mistral_api_key'))
 
-
 # Function to get chat response
 def get_mistral_response(user_input, history: list[tuple[str, str]], context):
     messages = [ChatMessage(role='user', content='you are helpful assistant answering question based on given context. provide a highly consie and precise answer, along with a citation from original context in the format of JSON.'),
                 ChatMessage(role='assistant', content='I understand.')] # system prompt has been baked into fine-tuned model. 
     
     context_with_history = context # placeholder
-    chat_response = client.chat(
+    response_after = client.chat(
         model='ft:open-mistral-7b:1c04df3c:20240629:7010a3c8',
         messages=[ChatMessage(role='user', content=f'''CONTEXT:{context_with_history}  QUESTION: {user_input}''')]
-    )
-    response = chat_response.choices[0].message.content
-    return response
-
-
-def get_zephyr_respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p):
-    messages = [{"role": "system", "content": system_message}]
-
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
+    ).choices[0].message.content
 
-    messages.append({"role": "user", "content": message})
-
-    response = ""
+    response_before = client.chat(
+        model='open-mistral-7b',
+        messages=[ChatMessage(role='user', content=f'''CONTEXT:{context_with_history}  QUESTION: {user_input}''')]
+    ).choices[0].message.content
+    
+    response = f"**Mistral before fine-tune**: <br> {response_before} <br> **Mistral after fine-tune**:<br> {response_after} <br>"
+    return response
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
 
-        response += token
-        yield response
 
 
 demo = gr.ChatInterface(
     get_mistral_response,
-    title='no-nonsense chatbot',
+    title='no-nonsense QA bot',
     description="This bot answers your question with grounded citation. Copy paste your contextual information in the box below.",
     additional_inputs=[
         gr.Textbox(value="", label="Answer will be based on the context input here", lines=10),