Update app.py

app.py

@@ -96,19 +96,24 @@ def generate(prompt, history, max_new_tokens=256): # temperature=0.95, top_p=0.9
 
     formatted_prompt = format_prompt(prompt, history)
 
-    # response = llm(formatted_prompt, **kwargs, stream=True)
+    response = llm(formatted_prompt, **kwargs, stream=True)
+    output = ""
+    for chunk in response:
+        output += chunk['choices'][0]['text']
+        yield output
+    return output
     # output = ""
     # for chunk in response:
     #     output += chunk.token.text
     #     yield output
     # return output
     
-    response = llm(formatted_prompt, **kwargs)
-    return response['choices'][0]['text']
+    # response = llm(formatted_prompt, **kwargs)
+    # return response['choices'][0]['text']
 
 chatbot = gr.Chatbot(height=500)
 with gr.Blocks(theme=gr.themes.Soft()) as demo:
-    gr.HTML("<center><h1>Google Gemma 7B IT</h1><center>")
+    gr.HTML("<center><h1>Fine-tuned Meta-Llama-3-8B</h1><center>")
     gr.ChatInterface(
         generate,
         chatbot=chatbot,