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

model_path = "RUSpam/spam_deberta_v4"
tokenizer = AutoTokenizer.from_pretrained(model_path)
model = AutoModelForSequenceClassification.from_pretrained(model_path)

def predict_spam(text):
    inputs = tokenizer(text, return_tensors="pt", truncation=True, max_length=256)
    with torch.no_grad():
        outputs = model(**inputs)
        logits = outputs.logits
        probabilities = torch.softmax(logits, dim=1)
        predicted_class = torch.argmax(logits, dim=1).item()
    
    spam_probability = probabilities[0][1].item()
    not_spam_probability = probabilities[0][0].item()
    
    result = "Спам" if predicted_class == 1 else "Не спам"
    
    return {
        "Результат": result,
        "Вероятность спама": f"{spam_probability:.2%}"
    }

# Создание интерфейса Gradio
iface = gr.Interface(
    fn=predict_spam,
    inputs=gr.Textbox(lines=5, label="Введите текст"),
    outputs=[
        gr.Label(label="Результат"),
        gr.Label(label="Вероятность спама")
    ],
    title="Определение спама в русскоязычных текстах"
)

iface.launch()