app.py

import gradio as gr
import torch
from bert_gru_classifier import BERTBiGRUClassifier
from bert_lstm_classifier import BERTBiLSTMClassifier
from transformers import AutoTokenizer

CLASS_MAP = {0: "Negative", 1: "Neutral", 2: "Positive" }

# Load tokenizer (pakai tokenizer yang sama untuk semua model)
tokenizer = AutoTokenizer.from_pretrained("LazarusNLP/NusaBERT-large")

# Load models
bigru_model = BERTBiGRUClassifier.from_pretrained("Amal17/NusaBERT-concate-BiGRU-NusaX-ace")
bigru_model.eval()

bigru_translate_model = BERTBiGRUClassifier.from_pretrained("Amal17/NusaBERT-concate-BiGRU-NusaTranslate-senti")
bigru_translate_model.eval()

bilstm_model = BERTBiLSTMClassifier.from_pretrained("Amal17/NusaBERT-concate-BiLSTM-NusaX-ace")
bilstm_model.eval()

bilstm_translate_model = BERTBiLSTMClassifier.from_pretrained("Amal17/NusaBERT-concate-BiLSTM-NusaTranslate-senti")
bilstm_translate_model.eval()



# Inference helper
def predict_with_model(model, text):
    inputs = tokenizer(
        text,
        padding="max_length",
        truncation=True,
        max_length=128,
        return_tensors="pt"
    )
    with torch.no_grad():
        outputs = model(**inputs)
        logits = outputs["logits"]
        probs = torch.softmax(logits, dim=1)
        pred = torch.argmax(probs, dim=1).item()
        confidence = probs[0][pred].item()
    return pred, confidence

# Gradio interface function
def compare_models(text):
    pred_a, conf_a = predict_with_model(bigru_model, text)
    pred_b, conf_b = predict_with_model(bilstm_model, text)

    pred_c, conf_c = predict_with_model(bigru_translate_model, text)
    pred_d, conf_d = predict_with_model(bilstm_translate_model, text)

    return (
        f"Class: {pred_a} ({CLASS_MAP[pred_a]}) with confidence: {conf_a:.4f}",
        f"Class: {pred_b} ({CLASS_MAP[pred_b]}) with confidence: {conf_b:.4f}",
        f"Class: {pred_c} ({CLASS_MAP[pred_c]}) with confidence: {conf_c:.4f}",
        f"Class: {pred_d} ({CLASS_MAP[pred_d]}) with confidence: {conf_d:.4f}",
    )

# Build Gradio UI
interface = gr.Interface(
    fn=compare_models,
    inputs=gr.Textbox(label="Input Text"),
    outputs=[
        gr.Textbox(label="NusaBERT-BiGRU-ace"),
        gr.Textbox(label="NusaBERT-BiLSTM-ace"),
        gr.Textbox(label="NusaBERT-BiGRU-translate"),
        gr.Textbox(label="NusaBERT-BiLSRM-translate"),
    ],
    title="Hybrid Model NusaBERT + RNN"
)

interface.launch()