Update app.py

app.py

@@ -1,33 +1,23 @@
 import streamlit as st
-from transformers import AutoModelForCausalLM, AutoTokenizer
+from transformers import pipeline
 import torch
 
-# Load DialoGPT model and tokenizer
-tokenizer = AutoTokenizer.from_pretrained("microsoft/DialoGPT-medium",padding_side='left')
-model = AutoModelForCausalLM.from_pretrained("microsoft/DialoGPT-medium")
+# Load TinyLlama chatbot pipeline
+pipe = pipeline("text-generation", model="TinyLlama/TinyLlama-1.1B-Chat-v1.0", torch_dtype=torch.bfloat16, device_map="auto")
 
 # Streamlit app header
-st.set_page_config(page_title="Conversational Model Demo", page_icon="🤖")
-st.header("Conversational Model Demo")
-
-# Initialize chat history
-chat_history_ids = None
+st.set_page_config(page_title="Chatbot Demo", page_icon="🤖")
+st.header("Chatbot Demo")
 
 # Input for user message
 user_message = st.text_input("You:", "")
 
 if st.button("Send"):
-    # Encode the new user input, add the eos_token and return a tensor in PyTorch
-    new_user_input_ids = tokenizer.encode(user_message + tokenizer.eos_token, return_tensors='pt')
-
-    # Append the new user input tokens to the chat history
-    bot_input_ids = torch.cat([chat_history_ids, new_user_input_ids], dim=-1) if chat_history_ids is not None else new_user_input_ids
-
-    # Generate a response while limiting the total chat history to 1000 tokens
-    chat_history_ids = model.generate(bot_input_ids, max_length=1000, pad_token_id=tokenizer.eos_token_id)
-
-    # Pretty print last output tokens from the bot
-    model_response = tokenizer.decode(chat_history_ids[:, bot_input_ids.shape[-1]:][0], skip_special_tokens=True)
+    # Use TinyLlama chatbot pipeline to generate a response
+    messages = [{"role": "system", "content": "You are a friendly chatbot who always responds in the style of a pirate"},
+                {"role": "user", "content": user_message}]
+    prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
+    response = pipe(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)[0]["generated_text"]
 
     # Display the model's response
-    st.text_area("Model:", model_response, height=100)
+    st.text_area("Model Response:", response, height=100)