Update: Initialization

app.py

@@ -1,57 +1,51 @@
 import gradio as gr
-from transformers import pipeline
+from huggingface_hub import InferenceClient
 import os
-from dotenv import load_dotenv
 
-load_dotenv()
-
-# Load model from Hugging Face Hub
-model_id = "VentureCircle/LaunchPad"
 token = os.getenv("huggingface_token")
-
-generator = pipeline(
-    "text-generation",
-    model=model_id,
-    tokenizer=model_id,
-    token=token
-)
-
-def respond(message, history, system_message, max_tokens, temperature, top_p):
-    # Combine chat history and message into a single prompt
-    chat_prompt = system_message + "\n\n"
+client = InferenceClient("HuggingFaceH4/zephyr-7b-beta", token=token)
+
+def respond(
+    message,
+    history: list[tuple[str, str]],
+    system_message,
+    max_tokens,
+    temperature,
+    top_p,
+):
+    messages = [{"role": "system", "content": system_message}]
     for user, assistant in history:
         if user:
-            chat_prompt += f"User: {user}\n"
+            messages.append({"role": "user", "content": user})
         if assistant:
-            chat_prompt += f"LaunchPad: {assistant}\n"
-    chat_prompt += f"User: {message}\nLaunchPad:"
-
-    output = generator(
-        chat_prompt,
-        max_new_tokens=max_tokens,
+            messages.append({"role": "assistant", "content": assistant})
+    messages.append({"role": "user", "content": message})
+
+    response = ""
+    for message in client.chat_completion(
+        messages,
+        max_tokens=max_tokens,
+        stream=True,
         temperature=temperature,
         top_p=top_p,
-        do_sample=True
-    )
-
-    # Extract response only (after "LaunchPad:")
-    response_text = output[0]["generated_text"].split("LaunchPad:")[-1].strip()
-    return response_text
-
-# System message configuration
-system_message = gr.State(
-    "I am Bilguun, a third year software engineering student at McGill University and the founder of VentureCircle. "
-    "VentureCircle is based in Montreal. "
-    "I created this chatbot, LaunchPad, to help aspiring entrepreneurs achieve their goals. "
-    "Your name is LaunchPad. "
-    "LaunchPad is a friendly AI assistant designed to provide useful and supportive guidance for starting and growing a startup."
-)
+    ):
+        token_piece = message.choices[0].delta.content
+        response += token_piece
+        yield response
 
-# Gradio interface
 demo = gr.ChatInterface(
-    fn=respond,
+    respond,
     additional_inputs=[
-        system_message,
+        gr.Textbox(
+            value=(
+                "I am Bilguun, a third year software engineering student at McGill University and the founder of VentureCircle. "
+                "VentureCircle is based in Montreal. "
+                "I created this chatbot, LaunchPad, to help aspiring entrepreneurs achieve their goals. "
+                "Your name is LaunchPad. "
+                "LaunchPad is a friendly AI assistant designed to provide useful and supportive guidance for starting and growing a startup."
+            ),
+            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(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"),