Pangeanic/Cantonese-English-Machine-Translation-Corpus-text

PangeanicYueEn - Cantonese English Parallel Corpus

PangeanicYueEn is a large-scale Cantonese-English parallel corpus designed for machine translation, multilingual large language model (LLM) training, cross-lingual NLP research, retrieval-augmented generation (RAG), bilingual embeddings, instruction tuning, and multilingual AI systems.

This release contains 150,000 Cantonese-English sentence pairs sampled from a larger corpus of approximately 8.05 million parallel sentence pairs. For the complete dataset, please contact Pangeanic

The dataset focuses specifically on written Cantonese (Yue Chinese) paired with English translations, providing valuable training data for one of the most underrepresented Chinese language varieties in modern AI systems.

---

Why Cantonese Matters for AI

Most multilingual AI systems are heavily trained on:

• English
• Mandarin Chinese
• Spanish
• French
• German

However, Cantonese (Yue Chinese) remains significantly underrepresented despite being spoken by tens of millions of people worldwide across:

• Hong Kong
• Macau
• Guangdong
• Southeast Asia
• North America
• Europe

As demand grows for multilingual AI, conversational assistants, translation systems, and regionally adapted LLMs, high-quality Cantonese datasets become increasingly important.

PangeanicYueEn helps address this gap by providing large-scale Cantonese-English parallel text suitable for modern AI development.

---

Dataset Overview

Attribute	Description
Language Pair	Cantonese (Yue) ↔ English
Dataset Type	Parallel Corpus
Modality	Text
Task	Machine Translation
Format	Parquet
Public Release Size	150,000 Sentence Pairs
Full Corpus Size	~8.05 Million Sentence Pairs
Script	Traditional Chinese
Domain	Mixed Web Content
Provider	Independent Dataset Release

---

Dataset Structure

The dataset contains two columns:

Column	Description
source	Cantonese sentence (Traditional Chinese)
target	English translation

Example:

source	target
培訓同教育都係必要嘅，以裝備人員掌握管理同運作新系統所需嘅技能。	Training and education are also necessary to equip personnel with the skills needed to manage and operate new systems effectively.
寫部落格係一個出色嘅兼職收入機會，因為佢讓你可以以自己嘅步伐同喺任何你選擇嘅地點工作。	Blogging presents an excellent opportunity for side income, as it allows you to work at your own pace and from any location you choose.
喺一個鄉村環境中，一名勞工同佢嘅配偶同孫子居住，反映出鄉村生活嘅簡樸。	In a rural setting, a laborer resided with his spouse and their grandchild, reflecting the simplicity of life in the countryside.

---

Example Usage

from datasets import load_dataset

dataset = load_dataset("Pangeanic/Cantonese-English-Machine-Translation-Corpus-text")

print(dataset["train"][0])

---

AI and LLM Use Cases

This dataset is suitable for:

• Cantonese machine translation
• English-Cantonese translation models
• multilingual LLM training
• continued pretraining
• supervised fine-tuning (SFT)
• retrieval-augmented generation (RAG)
• bilingual embedding models
• cross-lingual retrieval
• multilingual search systems
• conversational AI
• chatbot development
• language model evaluation
• synthetic data generation
• multilingual instruction tuning
• low-resource language research

---

Why This Dataset Is Useful

Large-Scale Cantonese Coverage

Cantonese remains one of the most underserved major languages in open AI training datasets.

This corpus provides substantial Cantonese-language coverage for researchers and developers working on multilingual AI systems.

Parallel Data for Translation

Aligned sentence pairs make the dataset suitable for:

• neural machine translation
• sequence-to-sequence models
• encoder-decoder architectures
• multilingual transformer training

LLM Adaptation

The dataset can be used for:

• multilingual LLM adaptation
• Cantonese language enhancement
• bilingual alignment
• continued pretraining
• instruction tuning

Cross-Lingual Applications

Developers can use PangeanicYueEn for:

• multilingual search
• semantic retrieval
• bilingual embeddings
• knowledge transfer
• multilingual RAG systems

---

Recommended Applications

Machine Translation

Train Cantonese ↔ English translation systems.

Multilingual Large Language Models

Improve Cantonese understanding and generation capabilities.

Conversational AI

Build Cantonese-capable assistants and chatbots.

Retrieval-Augmented Generation (RAG)

Enhance retrieval performance for Cantonese and English content.

Cross-Lingual Retrieval

Develop semantic search systems that operate across languages.

---

Dataset Characteristics

• Cantonese (Yue Chinese)
• Traditional Chinese script
• English translations
• Sentence-level alignment
• Large-scale bilingual data
• AI-ready format
• Suitable for modern NLP pipelines

---

Intended Research Areas

This dataset supports research in:

• machine translation
• multilingual NLP
• low-resource languages
• bilingual embeddings
• cross-lingual transfer learning
• language model adaptation
• multilingual retrieval
• multilingual RAG
• conversational AI
• Cantonese language technologies

---

Limitations

• This release is a sample of a larger corpus.
• The dataset does not include document-level metadata.
• The dataset is provided at sentence level.
• Domain distribution may vary across the corpus.

Users should evaluate suitability for their specific use cases and conduct additional filtering if required.

---

Citation

@dataset{pangeanic_cantonese_english_mt_2026,
  author = {{Pangeanic}},
  title = {Cantonese-English Machine Translation Corpus},
  year = {2026},
  publisher = {Pangeanic},
  url = {https://huggingface.co/datasets/Pangeanic/Cantonese-English-Machine-Translation-Corpus-text}
}

---

Keywords

Cantonese dataset, Yue dataset, Cantonese English corpus, Cantonese machine translation, bilingual dataset, parallel corpus, multilingual NLP, multilingual LLM, Cantonese AI, Chinese English translation, Cantonese chatbot dataset, Cantonese language model, machine translation corpus, bilingual AI training data, retrieval augmented generation, multilingual embeddings, low-resource language dataset, cross-lingual NLP, Cantonese NLP, Yue Chinese dataset.