| import gradio as gr | |
| import torch | |
| from transformers import PegasusForConditionalGeneration, PegasusTokenizer | |
| from sentence_splitter import SentenceSplitter, split_text_into_sentences | |
| 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=10000, | |
| return_tensors="pt").to(torch_device) | |
| translated = model.generate(**batch, 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 | |
| def get_response_from_text( | |
| context="I am a student at the University of Washington. I am taking a course called Data Science."): | |
| splitter = SentenceSplitter(language='en') | |
| sentence_list = splitter.split(context) | |
| 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 | |
| def greet(context): | |
| return get_response_from_text(context) | |
| iface = gr.Interface(fn=greet, inputs="text", outputs="text") | |
| iface.launch() |