Update app.py

app.py

@@ -1,48 +1,38 @@
 import streamlit as st
-import sys
-import subprocess
+from transformers import AutoModelForSequenceClassification, AutoTokenizer
+import torch
 
-def install(package):
-    subprocess.check_call([sys.executable, "-m", "pip", "install", package])
+# Load the model and tokenizer from Hugging Face
+model_name = "KevSun/Personality_LM"
+model = AutoModelForSequenceClassification.from_pretrained(model_name)
+tokenizer = AutoTokenizer.from_pretrained(model_name)
 
+# Streamlit app
 st.title("Personality Prediction App")
+st.write("Enter your text below to predict BigFive Personality traits:")
 
-# Check and install numpy library
-try:
-    import numpy as np
-except ImportError:
-    st.warning("The numpy library is not installed. Attempting to install it now...")
-    install('numpy')
-    st.experimental_rerun()
-
-# Check and install transformers library
-try:
-    from transformers import pipeline
-except ImportError:
-    st.warning("The transformers library is not installed. Attempting to install it now...")
-    install('transformers')
-    st.experimental_rerun()
-
-@st.cache_resource
-def load_model():
-    return pipeline("text-classification", model="KevSun/Personality_LM")
-
-model = load_model()
-
-st.write("Enter your text below to predict personality traits:")
-
+# Input text from user
 user_input = st.text_area("Your text here:")
 
 if st.button("Predict"):
     if user_input:
-        with st.spinner("Analyzing..."):
-            result = model(user_input)
+        # Tokenize input text
+        inputs = tokenizer(user_input, return_tensors="pt")
+        
+        # Get predictions from the model
+        with torch.no_grad():
+            outputs = model(**inputs)
+        
+        # Extract the predictions
+        predictions = torch.nn.functional.softmax(outputs.logits, dim=-1)
+        predictions = predictions[0].tolist()
         
-        st.subheader("Predicted personality traits:")
-        for trait in result:
-            st.write(f"- {trait['label']}: {trait['score']:.2f}")
+        # Display the predictions
+        labels = ["Extraversion", "Agreeableness", "Conscientiousness", "Neuroticism", "Openness"]
+        for label, score in zip(labels, predictions):
+            st.write(f"{label}: {score:.4f}")
     else:
-        st.warning("Please enter some text to analyze.")
+        st.write("Please enter your text.")
 
 st.info("Note: This is a demonstration and predictions may not be entirely accurate.")