app.py

from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
import gradio as gr
import torch

# Load model manually (bypasses pipeline issue)
model_name = "t5-small"

tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)

def summarize_text(text):
    if not text.strip():
        return "Please enter some text."

    input_text = "summarize: " + text

    inputs = tokenizer(
        input_text,
        return_tensors="pt",
        max_length=512,
        truncation=True
    )

    summary_ids = model.generate(
        inputs["input_ids"],
        max_length=120,
        min_length=30,
        length_penalty=2.0,
        num_beams=4,
        early_stopping=True
    )

    summary = tokenizer.decode(summary_ids[0], skip_special_tokens=True)
    return summary


demo = gr.Interface(
    fn=summarize_text,
    inputs=gr.Textbox(lines=10, placeholder="Enter long text here..."),
    outputs="text",
    title="📝 T5 Text Summarizer",
    description="Summarize long text using T5 (manual model loading)"
)

demo.launch()