app.py

import gradio as gr
import numpy as np
from tensorflow.keras.preprocessing.sequence import pad_sequences
from tensorflow.keras.models import load_model
import pickle

# Load the saved model
model = load_model('emotion_classifier_model.h5')

# Load the tokenizer (You need to save the tokenizer too)
with open('tokenizer.pickle', 'rb') as handle:
    tokenizer = pickle.load(handle)

# Define parameters for padding
max_length = 200
padding_type = 'post'
trunc_type = 'post'

# Define a function to predict emotions for a list of comments
def predict_emotions(comments):
    # Convert input text to sequences
    sequences = tokenizer.texts_to_sequences(comments)
    padded_sequences = pad_sequences(sequences, maxlen=max_length, padding=padding_type, truncating=trunc_type)

    # Predict emotions
    predictions = model.predict(padded_sequences)

    # List of emotion labels
    emotion_labels = ['admiration', 'amusement', 'anger', 'annoyance', 'approval', 'caring', 'confusion', 'curiosity', 
                      'desire', 'disappointment', 'disapproval', 'disgust', 'embarrassment', 'excitement', 'fear', 
                      'gratitude', 'grief', 'joy', 'love', 'nervousness', 'neutral', 'optimism', 'pride', 'realization', 
                      'relief', 'remorse', 'sadness', 'surprise']

    # Generate human-readable predictions
    result = []
    for prediction in predictions:
        emotion_dict = {emotion: prob for emotion, prob in zip(emotion_labels, prediction)}
        # Sort emotions by probability and get top 3
        top_emotions = sorted(emotion_dict.items(), key=lambda x: x[1], reverse=True)[:3]
        result.append({emotion: prob for emotion, prob in top_emotions})
    
    return result

# Create the Gradio interface
interface = gr.Interface(
    fn=predict_emotions, 
    inputs=gr.Textbox(label="Input Comment", lines=2, placeholder="Enter your comment here...", type="text"),
    outputs=gr.JSON(label="Predicted Emotions"),
    title="Reddit Emotion Classifier",
    description="Enter one or more comments and predict their emotion labels."
)

# Launch the app
interface.launch()