arxiv_class / tokenizer.py
kotstantinovskii's picture
Upload tokenizer.py
b6f331c
raw
history blame
226 Bytes
import streamlit as st
from transformers import DistilBertTokenizerFast
@st.cache(allow_output_mutation=True)
def get_tokenizer():
return DistilBertTokenizerFast.from_pretrained('distilbert-base-uncased', num_labels=172)