Spaces:
Sleeping
Sleeping
File size: 1,597 Bytes
c8766ff 3b64339 a9d6e8c c8766ff deb90aa e32e0a4 ab9ad6a e32e0a4 f306387 72c2b08 e32e0a4 01bb18b e32e0a4 ab9ad6a c8766ff deb90aa 9db86ab 12df622 374409a ab9ad6a 9db86ab ea0b887 f8bd591 ab9ad6a d7e74f5 889dd23 5781692 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 |
import gradio as gr
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, pipeline
device = "cpu"
tokenizer = AutoTokenizer.from_pretrained("humarin/chatgpt_paraphraser_on_T5_base")
model = AutoModelForSeq2SeqLM.from_pretrained("humarin/chatgpt_paraphraser_on_T5_base").to(device)
translator = pipeline("translation", model="facebook/nllb-200-distilled-600M")
def paraphrase(
question,
num_beams=5,
num_beam_groups=5,
num_return_sequences=1,
repetition_penalty=10.0,
diversity_penalty=3.0,
no_repeat_ngram_size=2,
temperature=0.7,
max_length=1024
):
input_ids = tokenizer(
f'paraphrase: {question}',
return_tensors="pt", padding="longest",
max_length=max_length,
truncation=True,
).input_ids.to(device)
outputs = model.generate(
input_ids, temperature=temperature, repetition_penalty=repetition_penalty,
num_return_sequences=num_return_sequences, no_repeat_ngram_size=no_repeat_ngram_size,
num_beams=num_beams, num_beam_groups=num_beam_groups,
max_length=max_length, diversity_penalty=diversity_penalty
)
res = tokenizer.batch_decode(outputs, skip_special_tokens=True)
return res
def translate(myinput):
myout = translator(myinput,src_lang="eng_Latn",tgt_lang="fra_Latn")
return myout
def predict(mytextInput):
out = translate(paraphrase(mytextInput))
return out
iface = gr.Interface(predict,
inputs="textbox",
outputs="text",
live=True,)
iface.launch(share=True)
|