app.py

import torch
from transformers import PegasusForConditionalGeneration, PegasusTokenizer

model_name = 'tuner007/pegasus_paraphrase'
torch_device = 'cuda' if torch.cuda.is_available() else 'cpu'
tokenizer = PegasusTokenizer.from_pretrained(model_name)
model = PegasusForConditionalGeneration.from_pretrained(model_name).to(torch_device)

def get_response(input_text,num_return_sequences):
  batch = tokenizer.prepare_seq2seq_batch([input_text],truncation=True,padding='longest',max_length=60, return_tensors="pt").to(torch_device)
  translated = model.generate(**batch,max_length=60,num_beams=10, num_return_sequences=num_return_sequences, temperature=1.5)
  tgt_text = tokenizer.batch_decode(translated, skip_special_tokens=True)
  return tgt_text

from sentence_splitter import SentenceSplitter, split_text_into_sentences

splitter = SentenceSplitter(language='en')

def paraphraze(text):
  sentence_list = splitter.split(text)
  paraphrase = []

  for i in sentence_list:
    a = get_response(i,1)
    paraphrase.append(a)
    paraphrase2 = [' '.join(x) for x in paraphrase]
    paraphrase3 = [' '.join(x for x in paraphrase2) ]
  paraphrased_text = str(paraphrase3).strip('[]').strip("'")
  return paraphrased_text

import gradio as gr
def summarize(text):

  paraphrased_text = paraphraze(text)
  return paraphrased_text
    
gr.Interface(fn=summarize, inputs=gr.inputs.Textbox(lines=7, placeholder="Enter text here"), outputs=[gr.outputs.Textbox(label="Paraphrased Text")], css="footer {visibility:hidden}", examples=[["This Api is the best quillbot api alternative with no words limit."
]]).launch(inline=False)