Update app.py

app.py

@@ -1,83 +1,56 @@
 import gradio as gr
+from openai import OpenAI
 import os
-import openai
-import tenacity
 import nest_asyncio
-import asyncio
 
 nest_asyncio.apply()
 
 ACCESS_TOKEN = os.getenv("HF_TOKEN")
-openai.api_key = ACCESS_TOKEN
 
-# Retry logic with tenacity for handling API rate limits
-@tenacity.retry(wait=tenacity.wait_exponential(multiplier=1, min=4, max=10), stop=tenacity.stop_after_attempt(5))
-async def respond(
+client = OpenAI(
+    base_url="https://api-inference.huggingface.co/v1/",
+    api_key=ACCESS_TOKEN,
+)
+
+def respond(
     message,
     system_message,
     max_tokens,
     temperature,
     top_p,
 ):
+    messages = [
+        {"role": "system", "content": system_message},
+        {"role": "user", "content": message}
+    ]
+    
     try:
-        print("Making request to OpenAI API...")
-        # Only use the system message and the current message for the response
-        messages = [{"role": "system", "content": system_message},
-                    {"role": "user", "content": message}]
-
-        response = await openai.Completion.create(
-            model="text-davinci-003",
-            prompt=system_message + "\n" + message,
+        print("Making request to API...")
+        response = client.chat.completions.create(
+            model="NousResearch/Hermes-3-Llama-3.1-8B",
+            messages=messages,
             max_tokens=max_tokens,
             temperature=temperature,
-            top_p=top_p,
+            top_p=top_p
         )
-
-        print("Received response from OpenAI API...")
-        response_text = response.choices[0].text
-        print("Response text:", response_text)
-        return response_text
-
-    except openai.error.APIError as e:
-        print("APIError:", e)
-        return "Error occurred. Please try again."
-
+        return response.choices[0].message.content
     except Exception as e:
-        print("Exception:", e)
-        return "Error occurred. Please try again."
-
-
-# Gradio function to handle user input and response generation without history
-def generate_response(message, system_message, max_tokens, temperature, top_p):
-    loop = asyncio.new_event_loop()
-    asyncio.set_event_loop(loop)
-    response = loop.run_until_complete(respond(message, system_message, max_tokens, temperature, top_p))
-    return response
-
-
-def launch_app():
-    try:
-        demo = gr.Blocks()
-        with demo:
-            gr.Markdown("# Chatbot")
-            message = gr.Textbox(label="Message")
-            system_message = gr.Textbox(label="System message")
-            max_tokens = gr.Slider(minimum=1, maximum=2048, value=2048, step=1, label="Max new tokens")
-            temperature = gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature")
-            top_p = gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-P")
-            response = gr.Text(label="Response")
-
-            # Use the generate_response function without history
-            gr.Button("Generate Response").click(
-                generate_response,
-                inputs=[message, system_message, max_tokens, temperature, top_p],
-                outputs=[response],
-                show_progress=False,
-            )
-        demo.launch(show_error=True)
-    except KeyError as e:
-        print("Error:", e)
-        print("Please try again.")
+        print(f"Error: {str(e)}")
+        return f"Error: {str(e)}"
+
+# Create simple Gradio interface without chat memory
+iface = gr.Interface(
+    fn=respond,
+    inputs=[
+        gr.Textbox(label="Message", lines=4),
+        gr.Textbox(label="System Message", value="You are a helpful assistant."),
+        gr.Slider(minimum=1, maximum=2048, value=512, label="Max Tokens"),
+        gr.Slider(minimum=0, maximum=1, value=0.7, label="Temperature"),
+        gr.Slider(minimum=0, maximum=1, value=0.9, label="Top P")
+    ],
+    outputs="text",
+    title="Hermes-3-Llama Chat (No Memory)"
+)
 
 if __name__ == "__main__":
-    launch_app()
+    iface.launch()