app.py

import tensorflow as tf
import gradio as gr
import numpy as np

# Load the custom model
model = tf.keras.models.load_model("sentiment.h5")

# Define a function to preprocess the input text
def preprocess_text(text):
    # Tokenize the text
    tokenizer = tf.keras.preprocessing.text.Tokenizer(num_words=10000)
    tokenizer.fit_on_texts(text)
    text = tokenizer.texts_to_sequences(text)
    # Pad the sequences to a fixed length
    max_len = 30
    text = tf.keras.preprocessing.sequence.pad_sequences(text, maxlen=max_len)
    return text

# Define a function to get the sentiment scores from the model
def get_sentiment_scores(text):
    # Preprocess the text
    text = preprocess_text(text)
    # Get the sentiment scores from the model
    scores = model.predict(text)
    # Postprocess the scores to obtain the sentiment label
    label = "Positive" if np.round(scores) == 1 else "Negative"
    return label

# Define the Gradio interface
interface = gr.Interface(
    fn=get_sentiment_scores,
    inputs=gr.inputs.Textbox(placeholder="Enter a positive or negative sentence here..."),
    outputs=gr.outputs.Textbox(label="Sentiment Label"),
    examples=[["This is wonderful!"], ["I hate this product."]]
)

# Launch the interface
interface.launch()