File size: 893 Bytes
7f8de92 bd063dc 1188df8 0f64e73 a2a9b24 b1b5dc2 0f64e73 b1b5dc2 0f64e73 854d0de a2a9b24 71af36e |
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 |
import streamlit as st
from transformers import pipeline
st.title("Milestone #2 v2")
text = st.text_input("write a statement")
import streamlit as st
options = ["zero-shot-classification", "cardiffnlp/twitter-roberta-base-offensive", "nill3"]
model = st.selectbox("Select an option", options)
##model = st.write("You selected:", selected_option)
con = st.button("submit")
if con:
if model == "zero-shot-classification":
classifier = pipeline(model)
res = classifier(text, candidate_labels= ["offensive"],)
print(res)
st.write(res)
if model == "cardiffnlp/twitter-roberta-base-offensive":
classifier = pipeline('text-classification', model='cardiffnlp/twitter-roberta-base-offensive', tokenizer='cardiffnlp/twitter-roberta-base-offensive')
result = offensive_classifier(text)
print(f" score {result[0]['score']*100:.2f}% confidence.")
|