app updated

app.py

@@ -1,29 +1,24 @@
 import streamlit as st
 from transformers import AutoTokenizer, AutoModelForSequenceClassification
 import torch
+import numpy as np
 
 
 # Define the function for sentiment analysis
 @st.cache_resource
-def predict_sentiment(text):
+def get_model():
    # Load the model and tokenizer
-    model = AutoModelForSequenceClassification.from_pretrained("MrDdz/mytuned_test_trainer-base-cased1")
     tokenizer = AutoTokenizer.from_pretrained("xlnet-base-cased")
+    model = AutoModelForSequenceClassification.from_pretrained("MrDdz/mytuned_test_trainer-base-cased1")
+tokenizer, model = get_model()
 
-    # Predict the sentiment.
-    prediction = (text)
-    sentiment = prediction[0]["label"]
-    score = prediction[0]["score"]
-
-    return sentiment, score
-
-# Setting the page configurations
-st.set_page_config(
-    page_title="Sentiment Analysis App",
-    page_icon=":smile:",
-    layout="wide",
-    initial_sidebar_state="auto",
-)
+# # Setting the page configurations
+# st.set_page_config(
+#     page_title="Sentiment Analysis App",
+#     page_icon=":smile:",
+#     layout="wide",
+#     initial_sidebar_state="auto",
+# )
 
 # Add description and title
 st.write("""
@@ -37,6 +32,13 @@ image = st.image("images.png", width=200)
 # Get user input
 text = st.text_input("Type here:")
 
+d = {
+
+    0:'Negative',
+    1:'Neutral',
+    2: 'Positive'
+     
+}
 # Define the CSS style for the app
 st.markdown(
 """
@@ -53,11 +55,18 @@ unsafe_allow_html=True
 )
 
 # Show sentiment output
+# if text:
+#     sentiment, score = predict_sentiment(text)
+#     if sentiment == "Positive":
+#         st.success(f"The sentiment is {sentiment} with a score of {score*100:.2f}%!")
+#     elif sentiment == "Negative":
+#         st.error(f"The sentiment is {sentiment} with a score of {score*100:.2f}%!")
+#     else:
+#         st.warning(f"The sentiment is {sentiment} with a score of {score*100:.2f}%!")
+
 if text:
-    sentiment, score = predict_sentiment(text)
-    if sentiment == "Positive":
-        st.success(f"The sentiment is {sentiment} with a score of {score*100:.2f}%!")
-    elif sentiment == "Negative":
-        st.error(f"The sentiment is {sentiment} with a score of {score*100:.2f}%!")
-    else:
-        st.warning(f"The sentiment is {sentiment} with a score of {score*100:.2f}%!")
+    text_sample = tokenizer([text], padding = True, Truncation = True, max_length = 512, return_tensors = 'pt' )
+    output = model(** text_sample)
+    st.write("Logits: ",output.logits)
+    y_pred = np.argmax(output.logits.detach().numpy(),axis =1)
+    st.write("Prediction :",d[y_pred[0]])