app.py

import gradio as gr
from transformers import AutoModelForSeq2SeqLM, AutoTokenizer
#from datasets import load_dataset
import torch

device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')

#SAVED_MODEL_PATH = 'bart_base_full_finetune_save'
SAVED_MODEL_PATH = './bart_base_full_finetune_save
model_name = "facebook/bart-base"
model = AutoModelForSeq2SeqLM.from_pretrained(SAVED_MODEL_PATH).to(device)
tokenizer = AutoTokenizer.from_pretrained(model_name)


def summarize(text):
    inputs = tokenizer(f"Summarize dialogue >>\n {text}", return_tensors="pt", max_length=1000, truncation=True, padding="max_length").to(device)
    summary_ids = model.generate(inputs.input_ids, num_beams=4, max_length=100, early_stopping=True)
    summary = [tokenizer.decode(g, skip_special_tokens=True, clean_up_tokenization_spaces=False) for g in summary_ids]
    return summary[0]

iface = gr.Interface(
    fn=summarize,
    inputs=gr.Textbox(lines=10, label="Input Dialogue"),
    outputs=gr.Textbox(label="Generated Summary")
)

iface.launch()