App 2.12

app.py

@@ -1,15 +1,10 @@
 import streamlit as st
 import pandas as pd
-import numpy as np
 from transformers import AutoTokenizer
 from transformers import (
     TFAutoModelForSequenceClassification as AutoModelForSequenceClassification,
 )
 
-def softmax(x):
-    e_x = np.exp(x - np.max(x))
-    return e_x / e_x.sum(axis=0)
-
 st.title("Detecting Toxic Tweets")
 
 demo = """Your words are like poison. They seep into my mind and make me feel worthless."""
@@ -36,11 +31,10 @@ def get_highest_toxicity_class(prediction):
     return model.config.id2label[max_index], prediction[max_index]
 
 input = tokenizer(text, return_tensors="tf")
-logits = model(input, return_dict=True).logits.numpy()[0]
-probabilities = softmax(logits)
+prediction = model(input, return_dict=True).logits.numpy()[0]
 
 if st.button("Submit", type="primary"):
-    label, probability = get_highest_toxicity_class(probabilities)
+    label, probability = get_highest_toxicity_class(prediction)
     
     tweet_portion = text[:50] + "..." if len(text) > 50 else text