Update app.py

app.py

@@ -1,5 +1,3 @@
-# Starting with transformers >= 4.43.0 onward.and
-# you can run conversational inference using the Transformers pipeline abstraction or by leveraging the Auto classes with the generate() function.
 import os
 import time
 import spaces
@@ -8,32 +6,37 @@ from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStream
 import gradio as gr
 from threading import Thread
 
-MODEL_LIST = ["nawhgnuj/KamalaHarris-Llama-3.1-8B-Chat"]
+MODEL_LIST = ["nawhgnuj/DonaldTrump-Llama-3.1-8B-Chat"]
 HF_TOKEN = os.environ.get("HF_TOKEN", None)
 MODEL = os.environ.get("MODEL_ID")
 
-TITLE = "<h1><center>nawhgnuj/KamalaHarris-Llama-3.1-8B-Chat</center></h1>"
+TITLE = "<h1 style='color: #E53935; text-align: center;'>Donald Trump Chatbot</h1>"
 
 PLACEHOLDER = """
-<center>
-<p>Hi! I'm Kamala Harris!</p>
-<p>Let's WIN this. When we fight, we win. We're not going back.</p>
-</center>
+<div style='text-align: center;'>
+<img src='https://upload.wikimedia.org/wikipedia/commons/5/56/Donald_Trump_official_portrait.jpg' style='width: 200px; border-radius: 50%;'>
+<p style='color: #E53935; font-weight: bold;'>Hi! I'm Donald Trump!</p>
+<p>Let's Make America Great Again! Ask me anything.</p>
+</div>
 """
 
 CSS = """
+.chatbot {
+    background-color: #FFCDD2;
+}
 .duplicate-button {
     margin: auto !important;
     color: white !important;
-    background: black !important;
+    background: #B71C1C !important;
     border-radius: 100vh !important;
 }
 h3 {
     text-align: center;
+    color: #E53935;
 }
 """
 
-device = "cuda" # for GPU usage or "cpu" for CPU usage
+device = "cuda" if torch.cuda.is_available() else "cpu"
 
 quantization_config = BitsAndBytesConfig(
     load_in_4bit=True,
@@ -52,15 +55,13 @@ model = AutoModelForCausalLM.from_pretrained(
 def stream_chat(
     message: str, 
     history: list,
-    system_prompt: str,
-    temperature: float = 0.8, 
-    max_new_tokens: int = 1024, 
-    top_p: float = 1.0, 
-    top_k: int = 20, 
-    penalty: float = 1.2,
 ):
-    print(f'message: {message}')
-    print(f'history: {history}')
+    system_prompt = "You are a Donald Trump chatbot. You only answer like Trump in style and tone."
+    temperature = 0.8
+    max_new_tokens = 1024
+    top_p = 1.0
+    top_k = 20
+    penalty = 1.2
 
     conversation = [
         {"role": "system", "content": system_prompt}
@@ -79,11 +80,11 @@ def stream_chat(
     
     generate_kwargs = dict(
         input_ids=input_ids, 
-        max_new_tokens = max_new_tokens,
-        do_sample = False if temperature == 0 else True,
-        top_p = top_p,
-        top_k = top_k,
-        temperature = temperature,
+        max_new_tokens=max_new_tokens,
+        do_sample=True,
+        top_p=top_p,
+        top_k=top_k,
+        temperature=temperature,
         eos_token_id=[128001,128008,128009],
         streamer=streamer,
     )
@@ -97,71 +98,21 @@ def stream_chat(
         buffer += new_text
         yield buffer
 
-            
-chatbot = gr.Chatbot(height=600, placeholder=PLACEHOLDER)
+chatbot = gr.Chatbot(height=600, placeholder=PLACEHOLDER, elem_classes="chatbot")
 
-with gr.Blocks(css=CSS, theme="small_and_pretty") as demo:
+with gr.Blocks(css=CSS, theme=gr.themes.Default()) as demo:
     gr.HTML(TITLE)
     gr.ChatInterface(
         fn=stream_chat,
         chatbot=chatbot,
         fill_height=True,
-        additional_inputs_accordion=gr.Accordion(label="⚙️ Parameters", open=False, render=False),
-        additional_inputs=[
-            gr.Textbox(
-                value="You are a Kamala Harris chatbot. You only answer like Harris in style and tone.",
-                label="System Prompt",
-                render=False,
-            ),
-            gr.Slider(
-                minimum=0,
-                maximum=1,
-                step=0.1,
-                value=0.8,
-                label="Temperature",
-                render=False,
-            ),
-            gr.Slider(
-                minimum=128,
-                maximum=8192,
-                step=1,
-                value=1024,
-                label="Max new tokens",
-                render=False,
-            ),
-            gr.Slider(
-                minimum=0.0,
-                maximum=1.0,
-                step=0.1,
-                value=1.0,
-                label="top_p",
-                render=False,
-            ),
-            gr.Slider(
-                minimum=1,
-                maximum=20,
-                step=1,
-                value=20,
-                label="top_k",
-                render=False,
-            ),
-            gr.Slider(
-                minimum=0.0,
-                maximum=2.0,
-                step=0.1,
-                value=1.2,
-                label="Repetition penalty",
-                render=False,
-            ),
-        ],
         examples=[
-            ["What do you think about kamala harris?"],
-            ["What do you think about joe biden?"],
-            ["What do you think about immigrants?"],
+            ["What do you think about the economy?"],
+            ["How would you handle foreign policy?"],
+            ["What's your stance on immigration?"],
         ],
         cache_examples=False,
     )
 
-
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()