sentiment analysis model's added

app.py

@@ -1,4 +1,34 @@
 import streamlit as st
+from transformers import pipeline, AutoTokenizer
 
-x = st.slider('Select a value')
-st.write(x, 'squared is', x * x)
+# Define a list of pretrained models
+models = {
+    "BERTweet": "finiteautomata/bertweet-base-sentiment-analysis",
+    "roBERTa": "cardiffnlp/twitter-roberta-base-sentiment",
+    "Distilbert": "bhadresh-savani/distilbert-base-uncased-emotion",
+    "BERT (BONUS!: Item Review)": "nlptown/bert-base-multilingual-uncased-sentiment",
+}
+
+# Display a selection box for the user to choose a model
+selected_model = st.selectbox("Select a pretrained model", list(models.keys()))
+
+# roBERTa specific label map
+roberta_label_map = {"LABEL_0": "negative", "LABEL_1": "neutral", "LABEL_2": "positive"}
+
+# Load the selected model and tokenizer
+model_name = models[selected_model]
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+sentiment_pipeline = pipeline("sentiment-analysis", model=model_name, tokenizer=tokenizer)
+
+# Get user input and perform sentiment analysis
+text_input = st.text_input("Enter text for sentiment analysis", "Anish is a very awesome and dedicated TA! 🤗")
+submit_btn = st.button("Submit")
+
+if submit_btn and text_input:
+    result = sentiment_pipeline(text_input)
+    if selected_model == "roBERTa":
+      st.write("Sentiment:", roberta_label_map[result[0]["label"]])
+      st.write("Score:", result[0]["score"])
+    else:
+      st.write("Sentiment:", result[0]["label"])
+      st.write("Score:", result[0]["score"])

requirements.txt

@@ -1,2 +1,4 @@
 streamlit
-numpy
+numpy
+transformers
+tensorflow