Create app.py

app.py

@@ -0,0 +1,51 @@
+import gradio as gr
+from transformers import pipeline, AutoTokenizer, AutoModelForSequenceClassification
+import shap
+import numpy as np
+
+# Load model and tokenizer
+def load_model():
+    model_name = "cardiffnlp/twitter-roberta-base-sentiment"
+    tokenizer = AutoTokenizer.from_pretrained(model_name)
+    model = AutoModelForSequenceClassification.from_pretrained(model_name)
+    sentiment_pipeline = pipeline("sentiment-analysis", model=model, tokenizer=tokenizer)
+    return sentiment_pipeline, tokenizer, model
+
+sentiment_pipeline, tokenizer, model = load_model()
+
+# Label mapping
+label_map = {"LABEL_0": "negative", "LABEL_1": "neutral", "LABEL_2": "positive"}
+
+# Gradio inference function
+def analyze_sentiment(text):
+    if not text.strip():
+        return "Please enter text.", "", None
+
+    # Get prediction
+    result = sentiment_pipeline(text)[0]
+    sentiment = label_map[result["label"]]
+    confidence = result["score"]
+
+    # Explain with SHAP
+    explainer = shap.Explainer(sentiment_pipeline)
+    shap_values = explainer([text])
+    explanation_html = shap.plots.text(shap_values[0], display=False)
+
+    sentiment_result = f"**Sentiment**: {sentiment.capitalize()}  \n**Confidence**: {confidence:.2f}"
+    return sentiment_result, shap_values.data[0], explanation_html
+
+# Gradio Interface
+iface = gr.Interface(
+    fn=analyze_sentiment,
+    inputs=gr.Textbox(label="Enter text", lines=4, placeholder="Type something here..."),
+    outputs=[
+        gr.Markdown(label="Sentiment Result"),
+        gr.Textbox(label="Tokenized Input"),
+        gr.HTML(label="SHAP Explanation"),
+    ],
+    title="Sentiment Analyzer with Explanation",
+    description="Enter a sentence or paragraph. This app will classify its sentiment as Positive, Negative, or Neutral and explain why using SHAP values.",
+)
+
+if __name__ == "__main__":
+    iface.launch()