app.py

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

import re



model_id = "google/gemma-1.1-2b-it"
dtype = torch.bfloat16

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    #device_map="cpu",
    torch_dtype=dtype,
)






st.title("💬 Chatbot")
st.caption("🚀 A streamlit chatbot powered by Google's Gemma")

# Initialize chat history
if 'messages' not in st.session_state:
    st.session_state['messages'] = [] #[{"role": "assistant", "content": "How can I help you?"}]

# Display chat messages from history on app rerun
for messasge in st.session_state.messages:
    st.chat_message(messasge["role"]).write(messasge["content"])

# React to user input
if prompt := st.chat_input():
    
     # Display user message in chat message container
    st.chat_message("user").write(prompt)
    # Add user message to chat history
    st.session_state.messages.append({"role": "user", "content": prompt})

    messages=st.session_state.messages

    
    text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)

    ##Get response to the message using client
    inputs = tokenizer.encode(text, add_special_tokens=False, return_tensors="pt")
    outputs = model.generate(input_ids=inputs, max_new_tokens=150)

    

    msg = tokenizer.decode(outputs[0]) #output[0]['generated_text']

    msg = re.sub(r'<.*?>', '', msg)
    
    
     # Display assistant response in chat message container
    st.chat_message("assistant").write(msg)

    # Add assistant response to chat history
    st.session_state.messages.append({"role": "assistant", "content": msg})