app.py

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

def choose_model(model_choice):
  if model_choice=="mt5-small":
    return "jannatul17/squad-bn-qgen-mt5-small-v1"
  elif model_choice=="mt5-base":
    return "Tahsin-Mayeesha/squad-bn-mt5-base2"
  else :
    return "jannatul17/squad-bn-qgen-banglat5-v1"


def generate__questions(model_choice,context,answer):
  model_name = choose_model(model_choice)
  model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
  tokenizer = AutoTokenizer.from_pretrained(model_name)  
  text='answer: '+answer + ' context: ' + context
  text_encoding = tokenizer.encode_plus(
      text,return_tensors="pt"
  )
  model.eval()
  generated_ids =  model.generate(
    input_ids=text_encoding['input_ids'],
    attention_mask=text_encoding['attention_mask'],
    max_length=64,
    num_beams=5,
    num_return_sequences=1
  )
  
  return tokenizer.decode(generated_ids[0],skip_special_tokens=True,clean_up_tokenization_spaces=True).replace('question: ',' ')

demo = gr.Interface(fn=generate__questions, inputs=[gr.Dropdown(label="Model", choices=["mt5-small","mt5-base","banglat5"],value="banglat5"),
                                                    gr.Textbox(label='Context'),
                                                    gr.Textbox(label='Answer')] ,
                                                    outputs=gr.Textbox(label='Question'),
                                                    title="Bangla Question Generation",
                                                    description="Get the Question from given Context and an Answer")
demo.launch()