goldfish-models / heb_hebr_1000mb

huggingface.co
Total runs: 35
24-hour runs: 0
7-day runs: 2
30-day runs: 25
Model's Last Updated: August 27 2024
text-generation

Introduction of heb_hebr_1000mb

Model Details of heb_hebr_1000mb

heb_hebr_1000mb

Goldfish is a suite of monolingual language models trained for 350 languages. This model is the Hebrew (Hebrew script) model trained on 1000MB of data, after accounting for an estimated byte premium of 1.36; content-matched text in Hebrew takes on average 1.36x 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: heb_hebr is an individual language code. It is not contained in any macrolanguage codes contained in Goldfish (for script hebr).

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): 1355.48MB
  • Training text data (byte premium scaled): 1000.005MB
  • Training tokens: 192904704 (x10 epochs)
  • Vocabulary size: 50000
  • Compute cost: 9.84445688807424e+17 FLOPs or ~93.1 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 heb_hebr_1000mb on huggingface.co

35
Total runs
0
24-hour runs
0
3-day runs
2
7-day runs
25
30-day runs

More Information About heb_hebr_1000mb huggingface.co Model

More heb_hebr_1000mb license Visit here:

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

heb_hebr_1000mb huggingface.co

heb_hebr_1000mb huggingface.co is an AI model on huggingface.co that provides heb_hebr_1000mb's model effect (), which can be used instantly with this goldfish-models heb_hebr_1000mb model. huggingface.co supports a free trial of the heb_hebr_1000mb model, and also provides paid use of the heb_hebr_1000mb. Support call heb_hebr_1000mb model through api, including Node.js, Python, http.

goldfish-models heb_hebr_1000mb online free

heb_hebr_1000mb huggingface.co is an online trial and call api platform, which integrates heb_hebr_1000mb's modeling effects, including api services, and provides a free online trial of heb_hebr_1000mb, you can try heb_hebr_1000mb online for free by clicking the link below.

goldfish-models heb_hebr_1000mb online free url in huggingface.co:

https://huggingface.co/goldfish-models/heb_hebr_1000mb

heb_hebr_1000mb install

heb_hebr_1000mb is an open source model from GitHub that offers a free installation service, and any user can find heb_hebr_1000mb on GitHub to install. At the same time, huggingface.co provides the effect of heb_hebr_1000mb install, users can directly use heb_hebr_1000mb installed effect in huggingface.co for debugging and trial. It also supports api for free installation.

heb_hebr_1000mb install url in huggingface.co:

https://huggingface.co/goldfish-models/heb_hebr_1000mb

Url of heb_hebr_1000mb

Provider of heb_hebr_1000mb huggingface.co

goldfish-models
ORGANIZATIONS

Other API from goldfish-models