CoolFace
Modelpublic

goldfish-models/hin_deva_1000mb

sourceHugging Faceapache-2.0updated 2y agoView on Hugging Face
0likes216downloads
Model Card

license: apache-2.0 language:

  • —hin datasets:
  • —oscar-corpus/OSCAR-2109 libraryname: transformers pipelinetag: text-generation tags:
  • —goldfish
  • —arxiv:2408.10441 ---

hindeva1000mb

Goldfish is a suite of monolingual language models trained for 350 languages. This model is the <b>Hindi</b> (Devanagari script) model trained on 1000MB of data, after accounting for an estimated byte premium of 2.37; content-matched text in Hindi takes on average 2.37x 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: This language is available in Goldfish with other scripts (writing systems). See: hin_latn.

Note: hindeva is an [individual language](https://iso639-3.sil.org/codetables/639/data) code. It is not contained in any macrolanguage codes contained in Goldfish (for script deva).

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. 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): 2370.16MB
  • —Training text data (byte premium scaled): 1000.005MB
  • —Training tokens: 228020736 (x10 epochs)
  • —Vocabulary size: 50000
  • —Compute cost: 1.163772239413248e+18 FLOPs or ~110.0 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},
  url={https://www.arxiv.org/abs/2408.10441},
}