goldfish-models / urd_arab_1000mb

huggingface.co
Total runs: 30
24-hour runs: 0
7-day runs: 4
30-day runs: 24
Model's Last Updated: August 27 2024
text-generation

Introduction of urd_arab_1000mb

Model Details of urd_arab_1000mb

urd_arab_1000mb

Goldfish is a suite of monolingual language models trained for 350 languages. This model is the Urdu (Arabic script) model trained on 1000MB of data, after accounting for an estimated byte premium of 1.71; content-matched text in Urdu takes on average 1.71x 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: urd_arab is an individual language code. It is not contained in any macrolanguage codes contained in Goldfish (for script arab).

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): 1707.95MB
  • Training text data (byte premium scaled): 1000.005MB
  • Training tokens: 247899648 (x10 epochs)
  • Vocabulary size: 50000
  • Compute cost: 1.265154592997376e+18 FLOPs or ~119.6 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 urd_arab_1000mb on huggingface.co

30
Total runs
0
24-hour runs
0
3-day runs
4
7-day runs
24
30-day runs

More Information About urd_arab_1000mb huggingface.co Model

More urd_arab_1000mb license Visit here:

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

urd_arab_1000mb huggingface.co

urd_arab_1000mb huggingface.co is an AI model on huggingface.co that provides urd_arab_1000mb's model effect (), which can be used instantly with this goldfish-models urd_arab_1000mb model. huggingface.co supports a free trial of the urd_arab_1000mb model, and also provides paid use of the urd_arab_1000mb. Support call urd_arab_1000mb model through api, including Node.js, Python, http.

goldfish-models urd_arab_1000mb online free

urd_arab_1000mb huggingface.co is an online trial and call api platform, which integrates urd_arab_1000mb's modeling effects, including api services, and provides a free online trial of urd_arab_1000mb, you can try urd_arab_1000mb online for free by clicking the link below.

goldfish-models urd_arab_1000mb online free url in huggingface.co:

https://huggingface.co/goldfish-models/urd_arab_1000mb

urd_arab_1000mb install

urd_arab_1000mb is an open source model from GitHub that offers a free installation service, and any user can find urd_arab_1000mb on GitHub to install. At the same time, huggingface.co provides the effect of urd_arab_1000mb install, users can directly use urd_arab_1000mb installed effect in huggingface.co for debugging and trial. It also supports api for free installation.

urd_arab_1000mb install url in huggingface.co:

https://huggingface.co/goldfish-models/urd_arab_1000mb

Url of urd_arab_1000mb

Provider of urd_arab_1000mb huggingface.co

goldfish-models
ORGANIZATIONS

Other API from goldfish-models