Update app.py

app.py

@@ -1,41 +1,47 @@
 # app.py
 
 import torch
-import streamlit as st
+import gradio as gr
 from transformers import AutoTokenizer, AutoModelForSequenceClassification
 
-# ---- Load model and tokenizer ----
-@st.cache_resource
-def load_model():
-    model_name = "taufiqdp/indonesian-sentiment"
-    tokenizer = AutoTokenizer.from_pretrained(model_name)
-    model = AutoModelForSequenceClassification.from_pretrained(model_name)
-    device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
-    model.to(device)
-    return tokenizer, model, device
-
-tokenizer, model, device = load_model()
-
-# ---- Streamlit App ----
-st.set_page_config(page_title="Indonesian Sentiment Analyzer", page_icon="🇮🇩", layout="centered")
-
-st.title("🇮🇩 Indonesian Sentiment Analyzer")
-st.markdown("### Analyze the sentiment of any Indonesian text — Positive, Neutral, or Negative.")
-
-text_input = st.text_area("Masukkan teks Bahasa Indonesia:", placeholder="Contoh: Pelayanan lama dan tidak ramah", height=120)
-
-if st.button("Analyze Sentiment"):
-    if not text_input.strip():
-        st.warning("⚠️ Mohon masukkan teks terlebih dahulu.")
-    else:
-        with st.spinner("Menganalisis..."):
-            inputs = tokenizer(text_input, return_tensors='pt', truncation=True, padding=True).to(device)
-            with torch.inference_mode():
-                logits = model(**inputs).logits
-                predicted_class = torch.argmax(logits, dim=1).item()
-
-            labels = ['Negatif', 'Netral', 'Positif']
-            sentiment = labels[predicted_class]
-
-            colors = {"Negatif": "red", "Netral": "orange", "Positif": "green"}
-            st.markdown(f"### 🔍 **Sentimen:** <span style='color:{colors[sentiment]};'>{sentiment}</span>", unsafe_allow_html=True)
+# --- Load model and tokenizer ---
+model_name = "taufiqdp/indonesian-sentiment"
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+model = AutoModelForSequenceClassification.from_pretrained(model_name)
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+model.to(device)
+
+# --- Prediction function ---
+def predict_sentiment(text):
+    if not text.strip():
+        return "⚠️ Mohon masukkan teks."
+
+    inputs = tokenizer(text, return_tensors='pt', truncation=True, padding=True).to(device)
+    with torch.inference_mode():
+        logits = model(**inputs).logits
+        pred_class = torch.argmax(logits, dim=1).item()
+
+    labels = ['Negatif', 'Netral', 'Positif']
+    sentiment = labels[pred_class]
+    return f"**Hasil Analisis Sentimen:** {sentiment}"
+
+# --- Build Gradio UI ---
+title = "🇮🇩 Indonesian Sentiment Analyzer"
+description = "Masukkan teks Bahasa Indonesia untuk mengetahui apakah sentimennya **Positif**, **Netral**, atau **Negatif**."
+
+interface = gr.Interface(
+    fn=predict_sentiment,
+    inputs=gr.Textbox(lines=4, placeholder="Contoh: Pelayanan cepat dan memuaskan"),
+    outputs="markdown",
+    title=title,
+    description=description,
+    theme="gradio/soft",
+    examples=[
+        ["Makanan di restoran ini sangat enak dan lezat."],
+        ["Pelayanan biasa saja, tidak buruk tapi tidak istimewa."],
+        ["Saya kecewa dengan produk ini."]
+    ]
+)
+
+if __name__ == "__main__":
+    interface.launch()