Spaces:
Runtime error
Runtime error
File size: 2,680 Bytes
afe86b6 |
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 61 62 63 64 65 66 67 68 69 |
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch
import nltk
nltk.download('punkt')
from nltk.tokenize import sent_tokenize
import streamlit as st
def load_model(model_id):
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)
return tokenizer, model
model_id = "asi/gpt-fr-cased-small"
tokenizer_fr, model_fr = load_model(model_id)
model_id = "gpt2"
tokenizer_en, model_en = load_model(model_id)
model_id = "dbmdz/german-gpt2"
tokenizer_de, model_de = load_model(model_id)
with st.form(key='Form'):
text = st.text_area("Enter text here.")
option = st.selectbox('Select Language',('English', 'German', 'French'))
submitted = st.form_submit_button("Submit")
if submitted:
text = text.replace('\n', '')
with torch.no_grad():
if option == 'German':
encodings = tokenizer_de(text, return_tensors="pt")
input_ids = encodings.input_ids
target_ids = input_ids.clone()
loss = model_de(input_ids, labels=target_ids).loss
elif option == 'English':
encodings = tokenizer_en(text, return_tensors="pt")
input_ids = encodings.input_ids
target_ids = input_ids.clone()
loss = model_en(input_ids, labels=target_ids).loss
else:
encodings = tokenizer_fr(text, return_tensors="pt")
input_ids = encodings.input_ids
target_ids = input_ids.clone()
loss = model_fr(input_ids, labels=target_ids).loss
st.write("Entire Text")
st.write("Perplexity: ", round(float(torch.exp(loss)), 2))
for sentence in sent_tokenize(text):
st.write("________________________")
st.write(sentence)
with torch.no_grad():
if option == 'German':
encodings = tokenizer_de(sentence, return_tensors="pt")
input_ids = encodings.input_ids
target_ids = input_ids.clone()
loss = model_de(input_ids, labels=target_ids).loss
elif option == 'English':
encodings = tokenizer_en(sentence, return_tensors="pt")
input_ids = encodings.input_ids
target_ids = input_ids.clone()
loss = model_en(input_ids, labels=target_ids).loss
else:
encodings = tokenizer_fr(sentence, return_tensors="pt")
input_ids = encodings.input_ids
target_ids = input_ids.clone()
loss = model_fr(input_ids, labels=target_ids).loss
st.write("Perplexity: ", round(float(torch.exp(loss)), 2)) |