Update app.py

app.py

@@ -1,11 +1,18 @@
 from huggingface_hub import InferenceClient
 import gradio as gr
-from werkzeug.urls import url_unquote
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
 
-client = InferenceClient("HuggingFaceH4/zephyr-7b-alpha")
+# Initialize the DialoGPT tokenizer and model
+tokenizer = AutoTokenizer.from_pretrained("microsoft/DialoGPT-large", padding_side="left")
+model = AutoModelForCausalLM.from_pretrained("microsoft/DialoGPT-large")
+
+client = InferenceClient(
+    "HuggingFaceH4/zephyr-7b-alpha"
+)
 
 def format_prompt(message, history):
-    system = "\nYou are a helpful virtual assistant that answers user's questions in a friendly manner.</s>\n"
+    system = "\nYou are a helpful virtual assistant that answers user's questions with easy-to-understand words.</s>\n"
     prompt = ""
     for user_prompt, bot_response in history:
         prompt += f"\n{user_prompt}</s>\n"
@@ -13,7 +20,14 @@ def format_prompt(message, history):
     prompt += f"\n{message}</s>\n"
     return prompt
 
-def generate_bot_response(prompt, history, temperature=0.9, max_new_tokens=500, top_p=0.95, repetition_penalty=1.0):
+def generate(
+    prompt,
+    history,
+    temperature=0.9,
+    max_new_tokens=500,
+    top_p=0.95,
+    repetition_penalty=1.0,
+):
     temperature = float(temperature)
     if temperature < 1e-2:
         temperature = 1e-2
@@ -30,7 +44,13 @@ def generate_bot_response(prompt, history, temperature=0.9, max_new_tokens=500,
 
     formatted_prompt = format_prompt(prompt, history)
 
-    stream = client.text_generation(formatted_prompt, **generate_kwargs, stream=True, details=True, return_full_text=False)
+    stream = client.text_generation(
+        formatted_prompt,
+        **generate_kwargs,
+        stream=True,
+        details=True,
+        return_full_text=False,
+    )
     output = ""
 
     for response in stream:
@@ -38,25 +58,65 @@ def generate_bot_response(prompt, history, temperature=0.9, max_new_tokens=500,
         yield output
     return output
 
-# Define your agents and prompts here
-agents = [
-    {"role": "convincing", "prompt": "I'm here to assist you. Please feel comfortable sharing your information."},
-    {"role": "extract", "prompt": "Great! Please provide your name."},
-    {"role": "extract", "prompt": "Thank you! Now, please provide your email address."},
-    {"role": "smalltalk", "prompt": "I understand that sharing personal information can be sensitive. Let's talk about something else for a moment."},
+additional_inputs = [
+    gr.Slider(
+        label="Temperature",
+        value=0.9,
+        minimum=0.0,
+        maximum=1.0,
+        step=0.05,
+        interactive=True,
+        info="Higher values produce more diverse outputs",
+    ),
+    gr.Slider(
+        label="Max new tokens",
+        value=256,
+        minimum=0,
+        maximum=1048,
+        step=64,
+        interactive=True,
+        info="The maximum number of new tokens",
+    ),
+    gr.Slider(
+        label="Top-p (nucleus sampling)",
+        value=0.90,
+        minimum=0.0,
+        maximum=1,
+        step=0.05,
+        interactive=True,
+        info="Higher values sample more low-probability tokens",
+    ),
+    gr.Slider(
+        label="Repetition penalty",
+        value=1.2,
+        minimum=1.0,
+        maximum=2.0,
+        step=0.05,
+        interactive=True,
+        info="Penalize repeated tokens",
+    ),
 ]
 
-with gr.Blocks() as inf:
+css = """
+  #mkd {
+    height: 500px;
+    overflow: auto;
+    border: 1px solid #ccc;
+  }
+"""
+
+with gr.Blocks(css=css) as inf:
+    gr.HTML("<h1><center>DialoGPT-large<h1><center>")
+    gr.HTML(
+        "<h3><center>In this demo, you can chat with <a href='https://huggingface.co/microsoft/DialoGPT-large'>DialoGPT-large</a> model. 💬<h3><center>"
+    )
     gr.ChatInterface(
-        generate_bot_response,
-        inputs=["", []],
-        outputs=["", []],
-        parameters=[
-            gr.Slider(label="Temperature", value=0.9, minimum=0.0, maximum=1.0, step=0.05, interactive=True, info="Higher values produce more diverse outputs"),
-            gr.Slider(label="Max new tokens", value=256, minimum=0, maximum=1048, step=64, interactive=True, info="The maximum number of new tokens"),
-            gr.Slider(label="Top-p (nucleus sampling)", value=0.90, minimum=0.0, maximum=1, step=0.05, interactive=True, info="Higher values sample more low-probability tokens"),
-            gr.Slider(label="Repetition penalty", value=1.2, minimum=1.0, maximum=2.0, step=0.05, interactive=True, info="Penalize repeated tokens"),
+        generate,
+        additional_inputs=additional_inputs,
+        examples=[
+            ["Can a squirrel swim?"],
+            ["Write a poem about a squirrel."],
         ],
     )
 
-inf.queue().launch()
+inf.queue().launch(share=True)