Spaces:
Sleeping
Sleeping
File size: 2,520 Bytes
9de6c4e 39e40d8 e578822 9de6c4e 39e40d8 2cbde77 39e40d8 2cbde77 9de6c4e 2cbde77 9de6c4e 2cbde77 9de6c4e 2cbde77 9de6c4e 39e40d8 2cbde77 39e40d8 9de6c4e 2cbde77 39e40d8 9de6c4e 2cbde77 9de6c4e 39e40d8 2cbde77 9de6c4e 39e40d8 9de6c4e |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 |
import gradio as gr
import joblib
from transformers import pipeline
# Load model dan pipeline
model = joblib.load("ensemble_model.pkl")
vectorizer = joblib.load("vectorizer.pkl")
qa_pipe = pipeline("question-answering", model="Rifky/IndoBERT-QA")
ner_pipe = pipeline("ner", model="cahya/bert-base-indonesian-NER", aggregation_strategy="simple")
# --- Fungsi ---
def detect_hoax(text):
vec = vectorizer.transform([text])
result = model.predict(vec)[0]
if result == 1:
return "<div style='background-color:#e74c3c; color:white; padding:10px; border-radius:5px'>HOAX</div>"
else:
return "<div style='background-color:#27ae60; color:white; padding:10px; border-radius:5px'>BUKAN HOAX</div>"
def qa_manual(message, history, context):
if not context:
return history + [[message, "Mohon isi teks berita terlebih dahulu."]]
result = qa_pipe(question=message, context=context)
return history + [[message, result["answer"]]]
def ner(text):
entities = ner_pipe(text)
styled = ""
color_map = {
"PER": "#ffd1dc", "ORG": "#d1e0ff", "LOC": "#d1ffd1", "MISC": "#fdfd96"
}
for ent in entities:
color = color_map.get(ent["entity_group"], "#eee")
styled += f"<mark style='background-color:{color}; padding:2px; margin:2px'>{ent['word']} <small>({ent['entity_group']})</small></mark> "
return styled
# --- UI Gradio ---
with gr.Blocks() as demo:
gr.Markdown("## Hoax Detector App")
context_input = gr.Textbox(label="Teks Berita / Konteks", lines=5, placeholder="Masukkan teks berita di sini...")
with gr.Tab("Deteksi Hoaks"):
detect_btn = gr.Button("DETEKSI")
hoax_output = gr.HTML()
detect_btn.click(fn=detect_hoax, inputs=context_input, outputs=hoax_output)
with gr.Tab("QA"):
#gr.Markdown("### Tanya Jawab Berdasarkan Teks Berita")
qa_question = gr.Textbox(placeholder="Tulis pertanyaan...", label="Pertanyaan")
qa_btn = gr.Button("KIRIM")
qa_history = gr.Chatbot(label="Riwayat Tanya Jawab")
qa_state = gr.State([])
qa_btn.click(
fn=qa_manual,
inputs=[qa_question, qa_state, context_input],
outputs=[qa_history],
show_progress=False
).then(fn=lambda h: h, inputs=qa_history, outputs=qa_state)
with gr.Tab("NER"):
ner_btn = gr.Button("Ekstrak Entitas")
ner_result = gr.HTML()
ner_btn.click(fn=ner, inputs=context_input, outputs=ner_result)
demo.launch()
|