bar plot percentages and a text table

app.py

@@ -1,5 +1,6 @@
 import gradio as gr
 from transformers import pipeline
+import pandas as pd
 
 # Load the Hugging Face pipeline
 classifier = pipeline("text-classification", model="bhadresh-savani/distilbert-base-uncased-emotion")
@@ -14,11 +15,20 @@ def classify_emotion(text):
     # Initialize a dictionary with all emotions and a default score of 0
     emotion_scores = {emotion: 0.0 for emotion in ALL_EMOTIONS}
     
-    # Update the dictionary with the raw scores returned by the model
+    # Update the dictionary with the scores returned by the model
     for item in predictions:
         emotion_scores[item["label"]] = item["score"]
     
-    return emotion_scores
+    # Create a dataframe for the bar plot
+    df = pd.DataFrame.from_dict(emotion_scores, orient="index").reset_index()
+    df.columns = ["Emotion", "Score"]
+    
+    # Prepare a text-based table for display
+    table = "Emotion Scores:\n"
+    table += "----------------\n"
+    table += "\n".join([f"{emotion}: {score:.4f}" for emotion, score in emotion_scores.items()])
+    
+    return df, table
 
 # Create a custom Gradio interface with title, description, and examples
 gr.Interface(
@@ -28,9 +38,20 @@ gr.Interface(
         label="Input Text",
         lines=4
     ),
-    outputs=gr.Label(label="Emotion Raw Scores"),  # Update label to reflect raw scores
-    title="CMACHINES | motion Detection with DistilBERT (Raw Scores)",
-    description="This app uses the DistilBERT model fine-tuned for emotion detection. Enter a piece of text to analyze its emotional content with raw scores!",
+    outputs=[
+        gr.BarPlot(
+            x="Emotion",
+            y="Score",
+            label="Emotion Scores Bar Plot",
+            title="Emotion Probabilities",
+            color="#2563eb",  # Color for the bars
+            height=400,
+            vertical=True
+        ),
+        gr.Textbox(label="Emotion Scores Table")  # Text-based table output
+    ],
+    title="CMACHINES25 | Emotion Detection with DistilBERT",
+    description="This app uses the DistilBERT model fine-tuned for emotion detection. Enter a piece of text to analyze its emotional content! Both a bar plot and a text table of the scores will be displayed.",
     examples=[
         "I like you. I love you. Sometimes I hate you.",
         "I'm really angry about what happened.",