| import streamlit as st | |
| from transformers import pipeline | |
| st.title("Milestone #2 offensive statement prediction") | |
| text = st.text_input("Enter a statement") | |
| options = ["zero-shot-classification", "cardiffnlp/twitter-roberta-base-offensive"] | |
| model = st.selectbox("Select a model", options) | |
| con = st.button("Submit") | |
| if con: | |
| if model == "zero-shot-classification": | |
| classifier = pipeline(model) | |
| res = classifier(text, candidate_labels=["offensive"]) | |
| label = res['labels'][0] | |
| score = res['scores'][0] | |
| st.write(f"Prediction: {label}, Score: {score*100}% chance") | |
| if model == "cardiffnlp/twitter-roberta-base-offensive": | |
| classifier = pipeline('text-classification', model='cardiffnlp/twitter-roberta-base-offensive', tokenizer='cardiffnlp/twitter-roberta-base-offensive') | |
| result = classifier(text) | |
| label = result[0]['label'] | |
| score = result[0]['score'] | |
| st.write(f"Prediction: {label}, Score: {score*100}% chance") | |