Add app.py with model predictor

app.py

@@ -0,0 +1,39 @@
+import gradio as gr
+import numpy as np
+
+from transformers import AutoTokenizer, TFAutoModelForSequenceClassification
+  
+tokenizer = AutoTokenizer.from_pretrained("uhhlt/bert-based-uncased-hatespeech-movies")
+
+model = TFAutoModelForSequenceClassification.from_pretrained("uhhlt/bert-based-uncased-hatespeech-movies")
+
+def make_prediction(text):
+    '''
+    This function takes a string as input and returns a prediction for the hate speech class.
+    Hate speech class labels are: Normal(0), Offensive(1), and Hate speech(2).
+
+    Parameters:
+        text (str): The text to be classified.
+
+    Returns:
+        str: The predicted class label.   
+    '''
+    input_ids = tokenizer.encode(text)
+    input_ids = np.array(input_ids)
+    input_ids = np.expand_dims(input_ids, axis=0)
+    prediction_arr = model.predict(input_ids)[0][0]
+
+    labels = ["Normal", "Offensive", "Hate Speech"]
+    prediction = labels[np.argmax(prediction_arr)]
+    return prediction
+
+
+
+iface = gr.Interface(
+  fn=make_prediction, 
+  inputs=gr.inputs.Textbox(lines=3, placeholder="Enter your text here..."), 
+  outputs="text",
+  title="Hate Speech Detector",
+  description="A model for detecting if a given text is an hate speech.",
+  )
+iface.launch()