goldfish-models / arb_arab_100mb

huggingface.co
Total runs: 99
24-hour runs: 0
7-day runs: 12
30-day runs: 72
Model's Last Updated: August 27 2024
text-generation

Introduction of arb_arab_100mb

Model Details of arb_arab_100mb

arb_arab_100mb

Goldfish is a suite of monolingual language models trained for 350 languages. This model is the Standard Arabic (Arabic script) model trained on 100MB of data, after accounting for an estimated byte premium of 1.47; content-matched text in Standard Arabic takes on average 1.47x 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: arb_arab is an individual language code. However, this model may also be useful for other Arabic dialects with less data.

All training and hyperparameter details are in our paper, Goldfish: Monolingual Language Models for 350 Languages (Chang et al., 2024) .

Training code and sample usage: https://github.com/tylerachang/goldfish

Sample usage also in this Google Colab: link

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. Details for this model specifically:

  • Architecture: gpt2
  • Parameters: 124770816
  • Maximum sequence length: 512 tokens
  • Training text data (raw): 146.50MB
  • Training text data (byte premium scaled): 100.005MB
  • Training tokens: 19608064 (x10 epochs)
  • Vocabulary size: 50000
  • Compute cost: 1.00043754504192e+17 FLOPs or ~9.5 NVIDIA A6000 GPU hours

Training datasets (percentages prior to deduplication):

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},
}

Runs of goldfish-models arb_arab_100mb on huggingface.co

99
Total runs
0
24-hour runs
6
3-day runs
12
7-day runs
72
30-day runs

More Information About arb_arab_100mb huggingface.co Model

More arb_arab_100mb license Visit here:

https://choosealicense.com/licenses/apache-2.0

arb_arab_100mb huggingface.co

arb_arab_100mb huggingface.co is an AI model on huggingface.co that provides arb_arab_100mb's model effect (), which can be used instantly with this goldfish-models arb_arab_100mb model. huggingface.co supports a free trial of the arb_arab_100mb model, and also provides paid use of the arb_arab_100mb. Support call arb_arab_100mb model through api, including Node.js, Python, http.

goldfish-models arb_arab_100mb online free

arb_arab_100mb huggingface.co is an online trial and call api platform, which integrates arb_arab_100mb's modeling effects, including api services, and provides a free online trial of arb_arab_100mb, you can try arb_arab_100mb online for free by clicking the link below.

goldfish-models arb_arab_100mb online free url in huggingface.co:

https://huggingface.co/goldfish-models/arb_arab_100mb

arb_arab_100mb install

arb_arab_100mb is an open source model from GitHub that offers a free installation service, and any user can find arb_arab_100mb on GitHub to install. At the same time, huggingface.co provides the effect of arb_arab_100mb install, users can directly use arb_arab_100mb installed effect in huggingface.co for debugging and trial. It also supports api for free installation.

arb_arab_100mb install url in huggingface.co:

https://huggingface.co/goldfish-models/arb_arab_100mb

Url of arb_arab_100mb

Provider of arb_arab_100mb huggingface.co

goldfish-models
ORGANIZATIONS

Other API from goldfish-models