emny/cdqa-indobert · Hugging Face

Details of cdqa-indobert (from their documentation)

[cdqa-indobert]://huggingface.co/emny/cdqa-indobert) is the Indonesian version of BERT model. We train the model using over 2M words, aggregated from three main sources:

Details of the downstream task (Q&A) - Dataset

The dataset contains information about Binus Admission. The data was collected from the Binus university’s admissions office for applicants who applied during the years 2020, 2021, and 2022. This dataset is provided to support researchers, especially in the field of natural language processing using the Indonesian language, and is not intended for commercialization purposes.

Dataset	Split	# samples
tanya-bion	train	10k
tanya-bion	eval	4k

Model Training

The model was trained on a Tesla A100 GPU and 83.5GB of RAM and 40GB of GPU RAM.

Results:

Metric	# Value
P	70.7
R	74.1
F1	72.1

Simple Usage

from transformers import pipeline

qa_pipeline = pipeline(
    "question-answering",
    model="emny/cdqa-indobert",
    tokenizer="emny/cdqa-indobert"
)

qa_pipeline({
    'context': """Jadwal Orientasi Binus Online area Jakarta, Bekasi, Palembang, Semarang, Malang pada 21-22 Mei 2024, Informasi jadwal pendaftaran dapat dilihat pada https://online.binus.ac.id/jadwal-pendaftaran/. Informasi dan Pendaftaran area Jakarta juga bisa didapat di Admisi Kampus Anggrek Universitas Binus dengan alamat Jl. Kebon Jeruk Raya No. 27, Kebon Jeruk, Jakarta Barat 11530. Informasi dan pendaftaran area Bekasi di Jl. Lingkar Bulevar Blok WA No.1 Summarecon Bekasi, Kelurahan Marga Mulya, Kecamatan Medan Satria, Bekasi 17142. Informasi dan pendaftaran area Palembang di Binus University Learning Community (Bulc) Palembang, Rukan Taman Harapan Indah (seberang Hotel Zuri Express), Jl. Letda A Rozak No. B3 dan B5, Palembang, Telepon (+62711) 5626222, Telepon seluler 081373381038, faksimil (+62 711) 5626666, dan email mbbolpalembang@binus.edu. Informasi dan Pendaftaran Admisi Kampus BINUS Semarang di BINUS Education Park (Gedung C), POJ Avenue Kav. 3, POJ City, Kel. Tawangsari, Kec. Semarang Barat, Kota Semarang. Informasi dan pendaftaran admisi Binus Malang Jl. Araya Mansion No 8-22, Araya Malang, telepon : (0341)3036969, telepon selular 081230 53288""",
    'question': "Kapan jadwal orientasi?"
})

output:

{
  'answer': '22 Mei 2024',
  'end': 99
  'score': 0.9673587083816528,
  'start': 88
}

Reference

[1]Emny Harna Yossy.