import tensorflow as tf
import gradio as gr
import pandas as pd
import os
from tensorflow.keras.layers import TextVectorization

df = pd.read_csv('train.csv')

X = df['comment_text']
y = df[df.columns[2:]].values

MAX_FEATURES = 200000 # number of words in the vocab

vectorizer = TextVectorization(max_tokens=MAX_FEATURES,
                              output_sequence_length=1800,
                              output_mode='int')

vectorizer.adapt(X.values)

model = tf.keras.models.load_model('hate_speech.h5')

def score_comment(comment):
    vectorized_comment = vectorizer([comment])
    results = model.predict(vectorized_comment)
    
    for idx, col in enumerate(df.columns[2:]):
        if results[0][idx]>0.5:
            return 'Hate Speech detected'
        
    return 'No hate speech detected'

interface = gr.Interface(fn=score_comment,
                        inputs=gr.Textbox(lines=2, placeholder='Comment to score'),
                        outputs='text')

interface.launch(share=True)