Update app.py

app.py

@@ -1,9 +1,10 @@
 import streamlit as st
 import pandas as pd
 import torch
-import asyncio
+from transformers import AutoTokenizer, AutoModelForSequenceClassification
+import random as r
 import gradio as gr
-from transformers import pipeline, AutoTokenizer, AutoModelForSequenceClassification
+import asyncio
 
 gr.Interface.load("models/APJ23/MultiHeaded_Sentiment_Analysis_Model").launch()
 
@@ -19,15 +20,14 @@ classes = {
     5: 'Insult',
     6: 'Identity Hate'
 }
-
 @st.cache(allow_output_mutation=True)
-def predict_toxicity(tweet, model, tokenizer):
+def predict_toxicity(tweet,model,tokenizer):
     inputs = tokenizer(tweet, return_tensors="pt", padding=True, truncation=True)
     outputs = model(**inputs)
     predicted_class = torch.argmax(outputs.logits, dim=1)
     predicted_prob = torch.softmax(outputs.logits, dim=1)[0][predicted_class].item()
     return classes[predicted_class], predicted_prob
-
+    
 def create_table(predictions):
     data = {'Tweet': [], 'Highest Toxicity Class': [], 'Probability': []}
     for tweet, prediction in predictions.items():
@@ -37,25 +37,23 @@ def create_table(predictions):
     df = pd.DataFrame(data)
     return df
 
-async def run_async_prediction(tweet, model, tokenizer):
-    loop = asyncio.get_event_loop()
-    prediction = await loop.run_in_executor(None, predict_toxicity, tweet, model, tokenizer)
-    return prediction
-
 st.title('Toxicity Prediction App')
-tweet_input = st.text_input('Enter a tweet to check for toxicity')
+tweet=st.text_input('Enter a tweet to check for toxicity')
 
-if st.button('Predict'):
-    predictions = {tweet_input: None}
+async def predict_toxicity_async(tweet, model, tokenizer):
     loop = asyncio.new_event_loop()
     asyncio.set_event_loop(loop)
-    prediction = loop.run_until_complete(run_async_prediction(tweet_input, model, tokenizer))
-    predictions[tweet_input] = prediction
+    result = await loop.run_until_complete(predict_toxicity(tweet, model, tokenizer))
     loop.close()
+    return result
+
+def predict_toxicity_sync(tweet, model, tokenizer):
+    return asyncio.run(predict_toxicity_async(tweet, model, tokenizer))
 
-    predicted_class_label, predicted_prob = prediction
+if st.button('Predict'):
+    predicted_class_label, predicted_prob = predict_toxicity_sync(tweet, model, tokenizer)
     prediction_text = f'Prediction: {predicted_class_label} ({predicted_prob:.2f})'
     st.write(prediction_text)
-    
+    predictions = {tweet: (predicted_class_label, predicted_prob)}
     table = create_table(predictions)
     st.table(table)