---
license: apache-2.0
language:
- sun
datasets:
- cis-lmu/Glot500
- allenai/c4
- legacy-datasets/wikipedia
- allenai/nllb
- allenai/MADLAD-400
- oscar-corpus/OSCAR-2109
library_name: transformers
pipeline_tag: text-generation
tags:
- goldfish
- arxiv:2408.10441
---

# sun_latn_full

Goldfish is a suite of monolingual language models trained for 350 languages.
This model is the <b>Sundanese</b> (Latin script) model trained on 577MB of data (all our data in the language), after accounting for an estimated byte premium of 1.10; content-matched text in Sundanese takes on average 1.10x as many UTF-8 bytes to encode as English.
The Goldfish models are trained primarily for comparability across languages and for low-resource languages; Goldfish performance for high-resource languages is not designed to be comparable with modern large language models (LLMs).

Note: sun_latn is an [individual language](https://iso639-3.sil.org/code_tables/639/data) code. It is not contained in any macrolanguage codes contained in Goldfish (for script latn).

All training and hyperparameter details are in our paper, [Goldfish: Monolingual Language Models for 350 Languages (Chang et al., 2024)](https://www.arxiv.org/abs/2408.10441).

Training code and sample usage: https://github.com/tylerachang/goldfish

Sample usage also in this Google Colab: [link](https://colab.research.google.com/drive/1rHFpnQsyXJ32ONwCosWZ7frjOYjbGCXG?usp=sharing)

## Model details:

To access all Goldfish model details programmatically, see https://github.com/tylerachang/goldfish/blob/main/model_details.json.
All models are trained with a [CLS] (same as [BOS]) token prepended, and a [SEP] (same as [EOS]) token separating sequences.
For best results, make sure that [CLS] is prepended to your input sequence (see sample usage linked above)!
Details for this model specifically:

* Architecture: gpt2
* Parameters: 124770816
* Maximum sequence length: 512 tokens
* Training text data (raw): 634.02MB
* Training text data (byte premium scaled): 577.965MB
* Training tokens: 142266368 (x10 epochs)
* Vocabulary size: 50000
* Compute cost: 7.26053214486528e+17 FLOPs or ~68.6 NVIDIA A6000 GPU hours

Training datasets (percentages prior to deduplication):
* 38.54837%: [Glot500](https://huggingface.co/datasets/cis-lmu/Glot500), including [CCNet](https://github.com/facebookresearch/cc_net), [Wortschatz Leipzig Data](https://wortschatz.uni-leipzig.de/en/download), [MC4](https://huggingface.co/datasets/allenai/c4), [OSCAR](https://oscar-project.org/), [W2C](https://lindat.mff.cuni.cz/repository/xmlui/handle/11858/00-097C-0000-0022-6133-9), [Wikipedia Hugging Face](https://huggingface.co/datasets/legacy-datasets/wikipedia)
* 32.02343%: [NLLB (CommonCrawl and ParaCrawl)](https://huggingface.co/datasets/allenai/nllb)
* 26.03343%: [MADLAD-400 (CommonCrawl)](https://huggingface.co/datasets/allenai/MADLAD-400)
* 2.55888%: [Wikipedia 2023/08](https://dumps.wikimedia.org/)
* 0.81131%: [IndoNLP](https://huggingface.co/indonlp)
* 0.02459%: [OSCAR 2021/09](https://huggingface.co/datasets/oscar-corpus/OSCAR-2109)


## Citation

If you use this model, please cite:

```
@article{chang-etal-2024-goldfish,
  title={Goldfish: Monolingual Language Models for 350 Languages},
  author={Chang, Tyler A. and Arnett, Catherine and Tu, Zhuowen and Bergen, Benjamin K.},
  journal={Preprint},
  year={2024},
  url={https://www.arxiv.org/abs/2408.10441},
}
```