Goldfish is a suite of monolingual language models trained for 350 languages.
This model is the
Komi
(Cyrillic script) model trained on 10MB of data, after accounting for an estimated byte premium of 1.61; content-matched text in Komi takes on average 1.61x 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: kom_cyrl is a
macrolanguage
code. Individual language code kpv_cyrl (Komi-Zyrian) is included in Goldfish, although with less data.
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: 39087104
Maximum sequence length: 512 tokens
Training text data (raw): 16.14MB
Training text data (byte premium scaled): 10.005MB
Training tokens: 2624512 (x10 epochs)
Vocabulary size: 50000
Compute cost: 1984941690716160.0 FLOPs or ~0.2 NVIDIA A6000 GPU hours
Training datasets (percentages prior to deduplication):
@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 kom_cyrl_10mb on huggingface.co
19
Total runs
0
24-hour runs
1
3-day runs
1
7-day runs
-13
30-day runs
More Information About kom_cyrl_10mb huggingface.co Model
kom_cyrl_10mb huggingface.co is an AI model on huggingface.co that provides kom_cyrl_10mb's model effect (), which can be used instantly with this goldfish-models kom_cyrl_10mb model. huggingface.co supports a free trial of the kom_cyrl_10mb model, and also provides paid use of the kom_cyrl_10mb. Support call kom_cyrl_10mb model through api, including Node.js, Python, http.
kom_cyrl_10mb huggingface.co is an online trial and call api platform, which integrates kom_cyrl_10mb's modeling effects, including api services, and provides a free online trial of kom_cyrl_10mb, you can try kom_cyrl_10mb online for free by clicking the link below.
goldfish-models kom_cyrl_10mb online free url in huggingface.co:
kom_cyrl_10mb is an open source model from GitHub that offers a free installation service, and any user can find kom_cyrl_10mb on GitHub to install. At the same time, huggingface.co provides the effect of kom_cyrl_10mb install, users can directly use kom_cyrl_10mb installed effect in huggingface.co for debugging and trial. It also supports api for free installation.