File size: 3,484 Bytes
48d19c7 5fee9eb 48d19c7 7368d8e 48d19c7 5fee9eb 48d19c7 f8023dd 48d19c7 5fee9eb 48d19c7 5fee9eb 48d19c7 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 |
---
license: apache-2.0
language:
- bak
datasets:
- cis-lmu/Glot500
- legacy-datasets/wikipedia
- allenai/MADLAD-400
- allenai/nllb
- oscar-corpus/OSCAR-2109
library_name: transformers
pipeline_tag: text-generation
tags:
- goldfish
- arxiv:2408.10441
---
# bak_cyrl_full
Goldfish is a suite of monolingual language models trained for 350 languages.
This model is the <b>Bashkir</b> (Cyrillic script) model trained on 398MB of data (all our data in the language), after accounting for an estimated byte premium of 2.27; content-matched text in Bashkir takes on average 2.27x 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: This language is available in Goldfish with other scripts (writing systems). See: bak_latn.
Note: bak_cyrl 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 cyrl).
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): 905.04MB
* Training text data (byte premium scaled): 398.365MB
* Training tokens: 118369280 (x10 epochs)
* Vocabulary size: 50000
* Compute cost: 6.04131216850944e+17 FLOPs or ~57.1 NVIDIA A6000 GPU hours
Training datasets (percentages prior to deduplication):
* 29.93711%: [Glot500](https://huggingface.co/datasets/cis-lmu/Glot500), including [Wortschatz Leipzig Data](https://wortschatz.uni-leipzig.de/en/download), [OSCAR](https://oscar-project.org/), [TIL](https://github.com/turkic-interlingua/til-mt), [Wikipedia Hugging Face](https://huggingface.co/datasets/legacy-datasets/wikipedia), [WikiMatrix](https://github.com/facebookresearch/LASER/tree/main/tasks/WikiMatrix)
* 29.35888%: [MADLAD-400 (CommonCrawl)](https://huggingface.co/datasets/allenai/MADLAD-400)
* 16.95765%: [NLLB (CommonCrawl and ParaCrawl)](https://huggingface.co/datasets/allenai/nllb)
* 11.91103%: [Wikipedia 2023/08](https://dumps.wikimedia.org/)
* 7.93919%: [Languages of Russia](http://web-corpora.net/wsgi3/minorlangs/download)
* 3.89615%: [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},
}
```
|