app.py

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

# Load local model
tokenizer = AutoTokenizer.from_pretrained("vanilladucky/Friends_chatting_bot")
model = AutoModelForCausalLM.from_pretrained("vanilladucky/Friends_chatting_bot")


def chat(user_input, history=[]):
    conversation = ""
    for human, bot in history:
        conversation += f"User: {human}\nBot: {bot}\n"
    conversation += f"User: {user_input}\nBot:"

    inputs = tokenizer(conversation, return_tensors="pt").to(model.device)

    outputs = model.generate(
        **inputs,
        max_new_tokens=100,
        do_sample=True,
        temperature=0.7,
        top_p=0.9
    )

    response = tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:], skip_special_tokens=True).strip()
    history.append((user_input, response))
    return history, history, ""   # <-- return "" to clear OR user_input to keep text


with gr.Blocks() as demo:
    gr.Markdown("CodeScape Demo")
    chatbot = gr.Chatbot()
    msg = gr.Textbox(label="Your message")
    clear = gr.Button("Clear")

    state = gr.State([])

    msg.submit(chat, [msg, state], [chatbot, state, msg])  # also update msg
    clear.click(lambda: ([], []), None, [chatbot, state])

demo.launch()