app.py

import gradio as gr
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

tokenizer = AutoTokenizer.from_pretrained("microsoft/DialoGPT-medium")
model = AutoModelForCausalLM.from_pretrained("microsoft/DialoGPT-medium")
max_history = 10  # Maximum number of previous chat turns to include in the conversation history
chat_history_ids = None

def chatbot(user_input):
    global chat_history_ids

    # encode the new user input, add the eos_token and return a tensor in PyTorch
    new_user_input_ids = tokenizer.encode(user_input + 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 max_history tokens
    chat_history_ids = model.generate(bot_input_ids, max_length=1000, pad_token_id=tokenizer.eos_token_id)

    # decode and return the generated response
    response = tokenizer.decode(chat_history_ids[:, bot_input_ids.shape[-1]:][0], skip_special_tokens=True)
    return response

styles = {
    "textarea": "height: 200px; font-size: 18px;",
    "label": "font-size: 20px; font-weight: bold;",
    "output": "color: red; font-size: 18px;"
}

iface = gr.Interface(fn=chatbot, inputs="text", outputs="text", title="Osana Chat Friend", styles=styles)
iface.launch()