app.py

import gradio as gr
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
from peft import PeftModel, PeftConfig

config = PeftConfig.from_pretrained("zeyadusf/FlanT5Summarization-samsum")
base_model = AutoModelForSeq2SeqLM.from_pretrained("google/flan-t5-large")
model = PeftModel.from_pretrained(base_model, "zeyadusf/FlanT5Summarization-samsum")
tokenizer = AutoTokenizer.from_pretrained("zeyadusf/FlanT5Summarization-samsum")

# Define the summarization function
def summarize(text):
    inputs = tokenizer(text, return_tensors="pt", truncation=True)
    # Access the base model's generate method
    summary_ids = model.base_model.generate(inputs.input_ids, max_length=512, min_length=64, length_penalty=2.0, num_beams=4, early_stopping=True)
    return tokenizer.decode(summary_ids[0], skip_special_tokens=True)

# Define the Gradio interface
iface = gr.Interface(fn=summarize, inputs="text", outputs="text", title="Summarization with PEFT")
iface.launch()