goldfish-models commited on
Commit
ad601ee
1 Parent(s): 2515d18

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +4 -2
README.md CHANGED
@@ -13,7 +13,7 @@ library_name: transformers
13
  pipeline_tag: text-generation
14
  tags:
15
  - goldfish
16
-
17
  ---
18
 
19
  # kur_arab_10mb
@@ -26,7 +26,7 @@ Note: This language is available in Goldfish with other scripts (writing systems
26
 
27
  Note: kur_arab is a [macrolanguage](https://iso639-3.sil.org/code_tables/639/data) code. Individual language code ckb_arab (Central Kurdish) is included in Goldfish, although with less data.
28
 
29
- All training and hyperparameter details are in our paper, [Goldfish: Monolingual Language Models for 350 Languages (Chang et al., 2024)](https://github.com/tylerachang/goldfish/blob/main/goldfish_paper_20240815.pdf).
30
 
31
  Training code and sample usage: https://github.com/tylerachang/goldfish
32
 
@@ -36,6 +36,7 @@ Sample usage also in this Google Colab: [link](https://colab.research.google.com
36
 
37
  To access all Goldfish model details programmatically, see https://github.com/tylerachang/goldfish/blob/main/model_details.json.
38
  All models are trained with a [CLS] (same as [BOS]) token prepended, and a [SEP] (same as [EOS]) token separating sequences.
 
39
  Details for this model specifically:
40
 
41
  * Architecture: gpt2
@@ -66,5 +67,6 @@ If you use this model, please cite:
66
  author={Chang, Tyler A. and Arnett, Catherine and Tu, Zhuowen and Bergen, Benjamin K.},
67
  journal={Preprint},
68
  year={2024},
 
69
  }
70
  ```
 
13
  pipeline_tag: text-generation
14
  tags:
15
  - goldfish
16
+ - arxiv:2408.10441
17
  ---
18
 
19
  # kur_arab_10mb
 
26
 
27
  Note: kur_arab is a [macrolanguage](https://iso639-3.sil.org/code_tables/639/data) code. Individual language code ckb_arab (Central Kurdish) is included in Goldfish, although with less data.
28
 
29
+ All training and hyperparameter details are in our paper, [Goldfish: Monolingual Language Models for 350 Languages (Chang et al., 2024)](https://www.arxiv.org/abs/2408.10441).
30
 
31
  Training code and sample usage: https://github.com/tylerachang/goldfish
32
 
 
36
 
37
  To access all Goldfish model details programmatically, see https://github.com/tylerachang/goldfish/blob/main/model_details.json.
38
  All models are trained with a [CLS] (same as [BOS]) token prepended, and a [SEP] (same as [EOS]) token separating sequences.
39
+ For best results, make sure that [CLS] is prepended to your input sequence (see sample usage linked above)!
40
  Details for this model specifically:
41
 
42
  * Architecture: gpt2
 
67
  author={Chang, Tyler A. and Arnett, Catherine and Tu, Zhuowen and Bergen, Benjamin K.},
68
  journal={Preprint},
69
  year={2024},
70
+ url={https://www.arxiv.org/abs/2408.10441},
71
  }
72
  ```