app.py

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

model = AutoModelForCausalLM.from_pretrained("DAMO-NLP-MT/polylm-1.7b")
model = PeftModel.from_pretrained(model, "fadliaulawi/polylm-1.7b-finetuned")
tokenizer = AutoTokenizer.from_pretrained("DAMO-NLP-MT/polylm-1.7b",use_fast = False)



def user(message, history):
    return "", history + [[message, None]]


def bot(history,temperature, max_length, top_p,top_k):
    user_message = history[-1][0]
    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([torch.LongTensor([]), new_user_input_ids], dim=-1)

    # generate a response
    response = model.generate(
        input_ids=bot_input_ids,
        pad_token_id=tokenizer.eos_token_id,
        temperature = float(temperature),
        max_length=max_length,
        top_p=float(top_p),
        top_k=top_k,
        do_sample=True
    ).tolist()

    # convert the tokens to text, and then split the responses into lines
    response = tokenizer.decode(response[0]).split("<|endoftext|>")
    response = [
        (response[i], response[i + 1]) for i in range(0, len(response) - 1, 2)
    ]  # convert to tuples of list
    history[-1] = response[0]
    return history


with gr.Blocks() as demo:
    temperature = gr.Slider(0, 5, value=0.8, step=0.1, label='Temperature')
    max_length = gr.Slider(0, 8192, value=256, step=1, label='Max Length')
    top_p = gr.Slider(0, 1, value=0.8, step=0.1, label='Top P')
    top_k = gr.Slider(0, 50, value=50, step=1, label='Top K')

    chatbot = gr.Chatbot()
    msg = gr.Textbox()
    submit = gr.Button("Submit")
    clear = gr.Button("Clear")

    examples = gr.Examples(examples=["Dokter aku sakit flu dan pilek. Apa yang terjadi denganku?"],inputs=[msg])

    #submit.click(bot,[msg,chatbot,temperature, max_length, top_p,top_k],chatbot)
    submit.click(user, [msg, chatbot], [msg, chatbot], queue=False).then(
        bot, [chatbot,temperature,max_length,top_p,top_k], chatbot
    )
    clear.click(lambda: None, None, chatbot, queue=False)

demo.queue().launch()