Hugging Face's logo

goldfish-models
/
msa_latn_1000mb

Text Generation
Transformers
PyTorch
Malay (macrolanguage)
may
gpt2
goldfish
text-generation-inference
Inference Endpoints
Model card Files Community
msa_latn_1000mb / README.md
goldfish-models's picture
goldfish-models
Upload README.md with huggingface_hub
a718348 verified
preview code
|
raw
history blame contribute delete
3.15 kB
metadata
license: apache-2.0
language:
  - msa
  - may
datasets:
  - allenai/MADLAD-400
  - cis-lmu/Glot500
  - oscar-corpus/OSCAR-2109
library_name: transformers
pipeline_tag: text-generation
tags:
  - goldfish
  - arxiv:2408.10441

msa_latn_1000mb

Goldfish is a suite of monolingual language models trained for 350 languages. This model is the Malay (Latin script) model trained on 1000MB of data, after accounting for an estimated byte premium of 1.29; content-matched text in Malay takes on average 1.29x 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: msa_latn is a macrolanguage code. Individual language codes meo_latn (Kedah Malay), bjn_latn (Banjar), min_latn (Minangkabau), ind_latn (Indonesian), and zsm_latn (Standard Malay)zsm_latn (Standard Malay), are included in Goldfish, although with less data.

All training and hyperparameter details are in our paper, Goldfish: Monolingual Language Models for 350 Languages (Chang et al., 2024).

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

Sample usage also in this Google Colab: link

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): 1285.71MB
  • Training text data (byte premium scaled): 1000.005MB
  • Training tokens: 236371456 (x10 epochs)
  • Vocabulary size: 50000
  • Compute cost: 1.206355524452352e+18 FLOPs or ~114.1 NVIDIA A6000 GPU hours

Training datasets (percentages prior to deduplication):

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},
}