import streamlit as st
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

# Load the model and tokenizer
model_id = "1bitLLM/bitnet_b1_58-3B"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

# Function to generate responses based on user messages
def generate_response(messages):
    input_ids = tokenizer.encode(messages, return_tensors="pt").to(model.device)
    outputs = model.generate(input_ids, max_length=100, pad_token_id=tokenizer.eos_token_id)
    generated_response = tokenizer.decode(outputs[0], skip_special_tokens=True)
    return generated_response

# Streamlit app
st.title("BitNet Chatbot")
messages = []

user_input = st.text_input("You:", "")
if st.button("Send"):
    if user_input:
        messages.append(user_input)
        bot_response = generate_response(messages)
        messages.append(bot_response)
    else:
        st.warning("Please enter a message.")

# Display conversation
for i, message in enumerate(messages):
    if i % 2 == 0:
        st.text_input("You:", value=message, disabled=True)
    else:
        st.text_area("BitNet:", value=message, disabled=True)