Update app.py

app.py

@@ -20,9 +20,18 @@ if not api_key:
 # Configure logging
 logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
 
-# Define constants
-PREFIX = "Task started at {date_time_str}. Purpose: {purpose}"
-TASK_PROMPT = "Current task: {task}. History:\n{history}"
+client = InferenceClient("mistralai/Mixtral-8x7B-Instruct-v0.1")
+
+VERBOSE = True
+MAX_HISTORY = 125
+
+def format_prompt(message, history):
+    prompt = "<s>"
+    for user_prompt, bot_response in history:
+        prompt += f"[INST] {user_prompt} [/INST]"
+        prompt += f" {bot_response}</s> "
+    prompt += f"[INST] {message} [/INST]"
+    return prompt
 
 # Define current date/time
 date_time_str = datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S")
@@ -110,8 +119,6 @@ def handle_input(task, storage_location, url1, url2, scrape_interval, content_ty
         return "You can chat with the friendly chatbot below."
 
 # Define the chat response function
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta", token=api_key)
-
 def respond(
     message,
     history,
@@ -126,33 +133,22 @@ def respond(
     scrape_interval,
     content_type
 ):
-    messages = [{"role": "system", "content": system_message}]
+    global client
 
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
+    # Format the prompt
+    prompt = format_prompt(message, history)
 
-    messages.append({"role": "user", "content": message})
+    # Generate response using the InferenceClient
+    response = client.generate(prompt, max_tokens=max_tokens, temperature=temperature, top_p=top_p)
 
-    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
+    # Append the response to history
+    history.append((message, response))
 
     # Handle the selected task
     handle_input(task, storage_location, url1, url2, scrape_interval, content_type)
 
+    return response
+
 # Create Gradio interface
 demo = gr.ChatInterface(
     respond,