Cendol Collection
IndoNLPCendol Collection is the instruction-tuning data IndoNLP used to train the Cendol family of language models for Indonesian and the country's local languages. Version 1 holds task-specific prompts built from NLP datasets for sentiment analysis, translation, summarization and question answering. Version 2 adds general-knowledge and human-centric prompts used for the Cendol-Chat models. Both are described in the Cendol paper.
Both dataset cards are copies of the Cendol model card and say little about how the prompts were assembled; the description relies on the model card's account of the two training stages.
Openness
4 medium confidence- license
- apache-2.0 (card metadata on both repos; the card text applies Apache 2.0 to 'use of Cendol')
- access
- public (both repos ungated)
- dataset_card
- partial (cards reproduce the model card and name the subset but do not describe sources or composition)
Both versions download without a gate under Apache-2.0. Their cards describe the Cendol models rather than the prompts, so how the data was assembled has to be read from the paper.
- https://huggingface.co/api/datasets/indonlp/cendol_collection_v2 recorded 2026-09-24
gated: false
- https://huggingface.co/datasets/indonlp/cendol_collection_v1/raw/main/README.md recorded 2026-09-24
license: apache-2.0; "This is the repository for the NusaT2T v2 - Task-Specific Prompts"
- https://huggingface.co/datasets/indonlp/cendol_collection_v2/raw/main/README.md recorded 2026-09-24
license: apache-2.0; "This is the repository for the NusaT2T v2 - General and Human Centric Prompts"
- https://raw.githubusercontent.com/IndoNLP/cendol/main/LICENSE recorded 2026-09-24
Apache License, Version 2.0 (repository LICENSE)
Adoption
1 high confidenceAdoption is Hugging Face downloads summed over the v1 and v2 repositories. Fine-tunes that reuse a local copy are not counted.
- https://huggingface.co/api/datasets/indonlp/cendol_collection_v1 recorded 2026-09-24
526 downloads in the trailing 30 days for indonlp/cendol_collection_v1
- https://huggingface.co/api/datasets/indonlp/cendol_collection_v2 recorded 2026-09-24
203 downloads in the trailing 30 days for indonlp/cendol_collection_v2
Capability
4 medium confidenceThe collection trained IndoNLP's Cendol Instruct and Chat models, from 300 million to 13 billion parameters, and the Cendol paper documents it. Like WangchanThaiInstruct it has a paper and named models trained on it, but its task prompts are rebuilt from existing labeled datasets and its cards do not describe the contents.
- https://arxiv.org/abs/2404.06138 recorded 2026-11-11
"we introduce Cendol, a collection of Indonesian LLMs encompassing both decoder-only and encoder-decoder architectures"
- https://arxiv.org/pdf/2404.06138 recorded 2026-11-11
Paper: "In total, we cover 187 safety prompts"; "In total, we cover 125 identity prompts"; "we add an Indonesian machine-translated dataset from Databricks-Dolly15k".
- https://datasets-server.huggingface.co/size?dataset=indonlp/cendol_collection_v1 recorded 2026-11-11
Release v1: "num_rows":30179364.
- https://datasets-server.huggingface.co/statistics?dataset=indonlp/cendol_collection_v2&config=default&split=train recorded 2026-11-11
Release v2 statistics, complete ("partial":false): "safety_prompt":93500, "identity_prompt":62500 and "dolly":75055 among the subsets.
- https://huggingface.co/api/datasets/indonlp/cendol_collection_v1 recorded 2026-11-11
size_categories:10M<n<100M
- https://huggingface.co/datasets/indonlp/cendol_collection_v2/raw/main/README.md recorded 2026-11-11
Release v2: "num_examples: 12810390".
- https://raw.githubusercontent.com/IndoNLP/cendol/main/README.md recorded 2026-11-11
Model table: "Cendol mT5-small Instruct](https://huggingface.co/indonlp/cendol-mt5-small-inst)|[Cendol Collection v1]"; "[Cendol mT5-small Chat](https://huggingface.co/indonlp/cendol-mt5-small-chat)|[Cendol Collection v2]".
Verified 2026-09-24