---
license: apache-2.0
language:
- yid
datasets:
- allenai/MADLAD-400
- allenai/nllb
- cis-lmu/Glot500
- sil-ai/bloom-lm
- oscar-corpus/OSCAR-2109
library_name: transformers
pipeline_tag: text-generation
tags:
- goldfish
- arxiv:2408.10441
---

# yid_hebr_full

Goldfish is a suite of monolingual language models trained for 350 languages.
This model is the <b>Yiddish</b> (Hebrew script) model trained on 446MB of data (all our data in the language), after accounting for an estimated byte premium of 1.55; content-matched text in Yiddish takes on average 1.55x 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: yid_hebr is a [macrolanguage](https://iso639-3.sil.org/code_tables/639/data) code. Individual language code ydd_hebr (Eastern Yiddish) is 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)](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): 689.62MB
* Training text data (byte premium scaled): 446.045MB
* Training tokens: 85695488 (x10 epochs)
* Vocabulary size: 50000
* Compute cost: 4.37368371019776e+17 FLOPs or ~41.4 NVIDIA A6000 GPU hours

Training datasets (percentages prior to deduplication):
* 58.13900%: [MADLAD-400 (CommonCrawl)](https://huggingface.co./datasets/allenai/MADLAD-400)
* 21.11646%: [NLLB (CommonCrawl and ParaCrawl)](https://huggingface.co./datasets/allenai/nllb)
* 9.12670%: [Glot500](https://huggingface.co./datasets/cis-lmu/Glot500), including [BLOOM](https://huggingface.co./datasets/sil-ai/bloom-lm), [CCNet](https://github.com/facebookresearch/cc_net), [Earthlings](https://publicdata.canterbury.ac.nz/Research/Geocorpus/CCGLU_v5.0/), [Wortschatz Leipzig Data](https://wortschatz.uni-leipzig.de/en/download), [OSCAR](https://oscar-project.org/), [W2C](https://lindat.mff.cuni.cz/repository/xmlui/handle/11858/00-097C-0000-0022-6133-9)
* 8.66901%: [OSCAR 2021/09](https://huggingface.co./datasets/oscar-corpus/OSCAR-2109)
* 2.94883%: [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},
}
```