Skip to content

Welcome to Lazarus NLP!

Lazarus NLP is a collective initiative to revive the dying languages of Indonesia through speech and language technology.


Revival of Languages, via OpenJourney


NusaBERT: Teaching IndoBERT to be multilingual and multicultural!

This project aims to extend the multilingual and multicultural capability of IndoBERT. We expanded the IndoBERT tokenizer on 12 new regional languages of Indonesia, and continued pre-training on a large-scale corpus consisting of the Indonesian language and 12 regional languages of Indonesia. Our models are highly competitive and robust on multilingual and multicultural benchmarks, such as IndoNLU, NusaX, and NusaWrites.

IndoT5: T5 Language Models for the Indonesian Language

IndoT5 is a T5-based language model trained specifically for the Indonesian language. With just 8 hours of training on a limited budget, we developed a competitive sequence-to-sequence, encoder-decode model capable of fine-tuning tasks such as summarization, chit-chat, and question-answering. Despite the limited training constraints, our model is competitive when evaluated on the IndoNLG (text generation) benchmark.

Indonesian Sentence Embedding Models

We trained open-source sentence embedding models for Indonesian, enabling applications such as information retrieval (useful for retrieval-augmented generation!) semantic text similarity, and zero-shot text classification. We leverage existing pre-trained Indonesian language models like IndoBERT and state-of-the-art unsupervised techniques and established sentence embedding benchmarks.

Indonesian Natural Language Inference Models

Open-source lightweight NLI models that are competitive with larger models on IndoNLI benchmark, with significantly less parameters. We applied knowledge distillation methods to small existing pre-trained language models like IndoBERT Lite. These models offer efficient solutions for tasks requiring natural language inference capabilities while minimizing computational resources such as cross-encoder-based semantic search.

Many-to-Many Multilingual Translation Models

Adapting mT5 to 45 languages of Indonesia, we developed a robust baseline model for multilingual translation for languages of Indonesia. This facilitates further fine-tuning for niche domains and low-resource languages, contributing to greater linguistic inclusivity. Our models are competitive with existing multilingual translation models on the NusaX benchmark.

Languages of Indonesia

Although the Indonesian language (Bahasa Indonesia) is the country's national language and lingua franca, the Indonesian people continue to use their region's native/indigenous language on a daily basis.

The nation is home to over 700 distinct languages, each with its own characteristics and origins. Some languages share similarities, but each language continues to evolve in its own ways over time and in the regions where the language is predominantly spoken.

Wikipedia1 gives a very nice overview of the spread and usage of several regional languages:

Language Number (millions) % of total population Branch Main areas where spoken
Javanese 84.3 32.28 Javanese throughout Java Island and several provinces in Sumatra and Kalimantan island.
Sundanese 42.0 16.08 Sundanese West Java, Banten, Jakarta
Madurese 13.6 5.21 Madurese Madura Island (East Java)
Minangkabau 5.5 2.11 Malayic West Sumatra, Riau, Jambi, Bengkulu, Jakarta
Buginese 5.0 1.91 South Sulawesi South Sulawesi
Palembang Malay 3.9 1.49 Malayic South Sumatra
Banjarese 3.5 1.34 Malayic South Kalimantan, East Kalimantan, Central Kalimantan
Acehnese 3.5 1.34 Chamic Aceh
Balinese 3.3 1.26 Bali-Sasak-Sumbawa Bali Island and Lombok Island
Betawi 2.7 1.03 Malay-based creole Jakarta

Tech 🤝 Languages

What is deeply concerning about these languages is that, although they may have millions of active speakers, they might still be prone to becoming endangered2. Furthermore, UNESCO classified 137 languages in Indonesia as either vulnerable, definitely endangered, severely endangered, or critically endangered3.

Despite the richness and diversity of these regional languages, there has very been minimal progress in the application of modern technologies to the many languages of Indonesia. It is only in recent years (~2020) that advancements began to appear, starting with the Indonesian language.

Through this project, we hope to see the same successes occur in other languages of Indonesia. In the long run, we also hope that through these technological advancements, we will be able to prevent these languages from becoming endangered and in turn, spark innovative work around these languages.


LazarusNLP is driven with love by:

  1. Wikipedia contributors. (2022, May 20). Languages of Indonesia. In Wikipedia, The Free Encyclopedia. Retrieved 07:55, May 21, 2022, from

  2. Abtahian, Maya & Cohn, Abigail. (2014). Can a language with millions of speakers be endangered?. Journal of the Southeast Asian Linguistics Society. 7. 64-75. 

  3. Moseley, Christopher, ed. (2010). Atlas of the World’s Languages in Danger. Memory of Peoples (3rd ed.). Paris: UNESCO Publishing. ISBN 978-92-3-104096-2.