File size: 3,589 Bytes
20ffcd0 000d73c 20ffcd0 000d73c 20ffcd0 1059af0 20ffcd0 000d73c 20ffcd0 000d73c 20ffcd0 |
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 |
---
license: apache-2.0
language:
- amh
datasets:
- cis-lmu/Glot500
- castorini/afriberta-corpus
- allenai/MADLAD-400
- allenai/nllb
- oscar-corpus/OSCAR-2109
library_name: transformers
pipeline_tag: text-generation
tags:
- goldfish
- arxiv:2408.10441
---
# amh_ethi_1000mb
Goldfish is a suite of monolingual language models trained for 350 languages.
This model is the <b>Amharic</b> (Ge'ez script) model trained on 1000MB of data, after accounting for an estimated byte premium of 1.72; content-matched text in Amharic takes on average 1.72x 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: amh_ethi 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 ethi).
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): 1720.92MB
* Training text data (byte premium scaled): 1000.005MB
* Training tokens: 211767808 (x10 epochs)
* Vocabulary size: 50000
* Compute cost: 1.080827696775168e+18 FLOPs or ~102.2 NVIDIA A6000 GPU hours
Training datasets (percentages prior to deduplication):
* 35.99036%: [Glot500](https://huggingface.co/datasets/cis-lmu/Glot500), including [AfriBERTa](https://huggingface.co/datasets/castorini/afriberta-corpus), [AfroMAFT](https://zenodo.org/record/6990611#.Y0-yU-xBw-Q), [CCNet](https://github.com/facebookresearch/cc_net), [Earthlings](https://publicdata.canterbury.ac.nz/Research/Geocorpus/CCGLU_v5.0/), [HornMT](https://github.com/asmelashteka/HornMT), [OSCAR](https://oscar-project.org/), [Parallel Corpora for Ethiopian Languages](https://github.com/AAUThematic4LT/Parallel-Corpora-for-Ethiopian-Languages), [TICO](https://tico-19.github.io/), [W2C](https://lindat.mff.cuni.cz/repository/xmlui/handle/11858/00-097C-0000-0022-6133-9)
* 33.89884%: [MADLAD-400 (CommonCrawl)](https://huggingface.co/datasets/allenai/MADLAD-400)
* 23.09942%: [NLLB (CommonCrawl and ParaCrawl)](https://huggingface.co/datasets/allenai/nllb)
* 6.48807%: [OSCAR 2021/09](https://huggingface.co/datasets/oscar-corpus/OSCAR-2109)
* 0.48971%: [Wikipedia 2023/08](https://dumps.wikimedia.org/)
* 0.03361%: [eBible](https://ebible.org/find/)
## 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},
}
```
|