metadata

license: apache-2.0
language:
  - lav
datasets:
  - oscar-corpus/OSCAR-2109
library_name: transformers
pipeline_tag: text-generation
tags:
  - goldfish

lav_latn_100mb

Goldfish is a suite of monolingual language models trained for 350 languages. This model is the Latvian (Latin script) model trained on 100MB of data, after accounting for an estimated byte premium of 1.29; content-matched text in Latvian 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: lav_latn is a macrolanguage code. Individual language codes lvs_latn (Standard Latvian) and ltg_latn (Latgalian) 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. Details for this model specifically:

Architecture: gpt2
Parameters: 124770816
Maximum sequence length: 512 tokens
Training text data (raw): 128.88MB
Training text data (byte premium scaled): 100.005MB
Training tokens: 24353792 (x10 epochs)
Vocabulary size: 50000
Compute cost: 1.24319613321216e+17 FLOPs or ~11.8 NVIDIA A6000 GPU hours

Training datasets (percentages prior to deduplication):

100.00000%: OSCAR 2021/09

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},
}