|
|
| --- |
| license: apache-2.0 |
| language: |
| - per |
| - fas |
| datasets: |
| - cis-lmu/Glot500 |
| library_name: transformers |
| pipeline_tag: text-generation |
| tags: |
| - goldfish |
| - arxiv:2408.10441 |
| --- |
| |
| # fas_arab_100mb |
|
|
| Goldfish is a suite of monolingual language models trained for 350 languages. |
| This model is the <b>Persian</b> (Arabic script) model trained on 100MB of data, after accounting for an estimated byte premium of 1.59; content-matched text in Persian takes on average 1.59x 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: fas_arab is a [macrolanguage](https://iso639-3.sil.org/code_tables/639/data) code. Individual language codes pes_arab (Iranian Persian) and prs_arab (Dari) 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)](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): 159.08MB |
| * Training text data (byte premium scaled): 100.005MB |
| * Training tokens: 24438272 (x10 epochs) |
| * Vocabulary size: 50000 |
| * Compute cost: 1.2475910651904e+17 FLOPs or ~11.8 NVIDIA A6000 GPU hours |
|
|
| Training datasets (percentages prior to deduplication): |
| * 100.00000%: [Glot500](https://huggingface.co/datasets/cis-lmu/Glot500), including [CCNet](https://github.com/facebookresearch/cc_net), [TICO](https://tico-19.github.io/), [W2C](https://lindat.mff.cuni.cz/repository/xmlui/handle/11858/00-097C-0000-0022-6133-9), [WikiMatrix](https://github.com/facebookresearch/LASER/tree/main/tasks/WikiMatrix) |
|
|
|
|
| ## 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}, |
| } |
| ``` |
|
|
