from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch
import gradio as gr

# Load pretrained hate-speech model
MODEL_NAME = "Hate-speech-CNERG/dehatebert-mono-english"
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
model = AutoModelForSequenceClassification.from_pretrained(MODEL_NAME)

def detect_hate(text):
    inputs = tokenizer(text, return_tensors="pt", truncation=True, padding=True)
    outputs = model(**inputs)
    probs = torch.nn.functional.softmax(outputs.logits, dim=-1)
    labels = ["non-hate", "hate"]
    result = {labels[i]: float(probs[0][i]) for i in range(len(labels))}
    return result

# Gradio interface
demo = gr.Interface(
    fn=detect_hate,
    inputs=gr.Textbox(label="Enter Text", placeholder="Type something..."),
    outputs=gr.Label(label="Prediction"),
    title="🧠 Hate Speech Detector",
    description="Classifies text as hate or non-hate using a fine-tuned BERT model.",
)

demo.launch()