File size: 3,021 Bytes
6a426f2 db506c6 6a426f2 db506c6 6a426f2 b2cf95b 6a426f2 db506c6 6a426f2 db506c6 6a426f2 |
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 |
---
license: apache-2.0
language:
- fur
datasets:
- breakend/nllb-multi-domain
- allenai/nllb
- cis-lmu/Glot500
- legacy-datasets/wikipedia
library_name: transformers
pipeline_tag: text-generation
tags:
- goldfish
- arxiv:2408.10441
---
# fur_latn_full
Goldfish is a suite of monolingual language models trained for 350 languages.
This model is the <b>Friulian</b> (Latin script) model trained on 20MB of data (all our data in the language), after accounting for an estimated byte premium of 1.07; content-matched text in Friulian takes on average 1.07x 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: fur_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): 22.11MB
* Training text data (byte premium scaled): 20.725MB
* Training tokens: 5487616 (x10 epochs)
* Vocabulary size: 50000
* Compute cost: 2.7999531565056e+16 FLOPs or ~2.6 NVIDIA A6000 GPU hours
Training datasets (percentages prior to deduplication):
* 72.56517%: [NLLB (CommonCrawl and ParaCrawl)](https://huggingface.co/datasets/allenai/nllb) and [NLLB Multi-Domain](https://huggingface.co/datasets/breakend/nllb-multi-domain)
* 15.82188%: [Glot500](https://huggingface.co/datasets/cis-lmu/Glot500), including [NLLB_seed](https://github.com/facebookresearch/flores/blob/main/nllb_seed/README.md), [Wikipedia Hugging Face](https://huggingface.co/datasets/legacy-datasets/wikipedia)
* 11.61295%: [Wikipedia 2023/08](https://dumps.wikimedia.org/)
## 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},
}
```
|