File size: 3,373 Bytes
5d0ad25
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a6e3f8c
5d0ad25
 
 
 
 
 
 
 
 
 
a6e3f8c
5d0ad25
 
 
 
 
 
 
6f1f5fe
5d0ad25
a6e3f8c
5d0ad25
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a6e3f8c
5d0ad25
 
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

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

# bod_tibt_100mb

Goldfish is a suite of monolingual language models trained for 350 languages.
This model is the <b>Tibetan</b> (Tibetan script) model trained on 100MB of data, after accounting for an estimated byte premium of 2.62; content-matched text in Tibetan takes on average 2.62x 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: bod_tibt 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 tibt).

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): 261.69MB
* Training text data (byte premium scaled): 100.005MB
* Training tokens: 17793536 (x10 epochs)
* Vocabulary size: 50000
* Compute cost: 9.0806558588928e+16 FLOPs or ~8.6 NVIDIA A6000 GPU hours

Training datasets (percentages prior to deduplication):
* 35.45938%: [OSCAR 2021/09](https://huggingface.co/datasets/oscar-corpus/OSCAR-2109)
* 27.32632%: [Glot500](https://huggingface.co/datasets/cis-lmu/Glot500), including [BLOOM](https://huggingface.co/datasets/sil-ai/bloom-lm), [Earthlings](https://publicdata.canterbury.ac.nz/Research/Geocorpus/CCGLU_v5.0/), [Wortschatz Leipzig Data](https://wortschatz.uni-leipzig.de/en/download), [MoT](https://github.com/bltlab/mot), [OSCAR](https://oscar-project.org/), [Wikipedia Hugging Face](https://huggingface.co/datasets/legacy-datasets/wikipedia)
* 23.94564%: [MADLAD-400 (CommonCrawl)](https://huggingface.co/datasets/allenai/MADLAD-400)
* 10.69915%: [Wikipedia 2023/08](https://dumps.wikimedia.org/)
* 1.63626%: [NLLB (CommonCrawl and ParaCrawl)](https://huggingface.co/datasets/allenai/nllb)
* 0.93326%: [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},
}
```