app.py

import gradio as gr
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM

# Load the model and tokenizer
model_name = "HafijulHoquenabid2/T5_flanlarge_phase_1"  # Replace with your model if needed
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)

# Define the function to generate answers
def generate_response(question, context):
    # Combine question and context
    input_text = f"question: {question} context: {context}"
    
    # Tokenize the input with attention mask
    inputs = tokenizer(
        input_text,
        max_length=512,
        padding="max_length",
        truncation=True,
        return_tensors="pt"
    )
    
    # Generate response
    outputs = model.generate(
        input_ids=inputs["input_ids"],
        attention_mask=inputs["attention_mask"],
        max_length=150,
        num_beams=4,
        early_stopping=True
    )
    
    # Decode and return the response
    answer = tokenizer.decode(outputs[0], skip_special_tokens=True)
    return answer

# Define Gradio interface
interface = gr.Interface(
    fn=generate_response,
    inputs=[
        gr.Textbox(label="Question", placeholder="Enter your question here..."),
        gr.Textbox(label="Context", placeholder="Provide the context for the question...")
    ],
    outputs="text",
    title="T5 Smart Door Lock QA",
    description="Ask domain-specific questions related to smart door locks and receive precise answers."
)

# Launch the app
if __name__ == "__main__":
    interface.launch()