goldfish-models commited on
Commit
15866c5
1 Parent(s): 8a8aa16

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +5 -3
README.md CHANGED
@@ -2,8 +2,8 @@
2
  ---
3
  license: apache-2.0
4
  language:
5
- - aym
6
  - ayr
 
7
  datasets:
8
  - breakend/nllb-multi-domain
9
  - allenai/nllb
@@ -11,7 +11,7 @@ library_name: transformers
11
  pipeline_tag: text-generation
12
  tags:
13
  - goldfish
14
-
15
  ---
16
 
17
  # ayr_latn_full
@@ -22,7 +22,7 @@ The Goldfish models are trained primarily for comparability across languages and
22
 
23
  Note: ayr_latn is an [individual language](https://iso639-3.sil.org/code_tables/639/data) code. Macrolanguage code aym_latn (Aymara) is included in Goldfish. Consider using that model depending on your use case.
24
 
25
- 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).
26
 
27
  Training code and sample usage: https://github.com/tylerachang/goldfish
28
 
@@ -32,6 +32,7 @@ Sample usage also in this Google Colab: [link](https://colab.research.google.com
32
 
33
  To access all Goldfish model details programmatically, see https://github.com/tylerachang/goldfish/blob/main/model_details.json.
34
  All models are trained with a [CLS] (same as [BOS]) token prepended, and a [SEP] (same as [EOS]) token separating sequences.
 
35
  Details for this model specifically:
36
 
37
  * Architecture: gpt2
@@ -57,5 +58,6 @@ If you use this model, please cite:
57
  author={Chang, Tyler A. and Arnett, Catherine and Tu, Zhuowen and Bergen, Benjamin K.},
58
  journal={Preprint},
59
  year={2024},
 
60
  }
61
  ```
 
2
  ---
3
  license: apache-2.0
4
  language:
 
5
  - ayr
6
+ - aym
7
  datasets:
8
  - breakend/nllb-multi-domain
9
  - allenai/nllb
 
11
  pipeline_tag: text-generation
12
  tags:
13
  - goldfish
14
+ - arxiv:2408.10441
15
  ---
16
 
17
  # ayr_latn_full
 
22
 
23
  Note: ayr_latn is an [individual language](https://iso639-3.sil.org/code_tables/639/data) code. Macrolanguage code aym_latn (Aymara) is included in Goldfish. Consider using that model depending on your use case.
24
 
25
+ 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).
26
 
27
  Training code and sample usage: https://github.com/tylerachang/goldfish
28
 
 
32
 
33
  To access all Goldfish model details programmatically, see https://github.com/tylerachang/goldfish/blob/main/model_details.json.
34
  All models are trained with a [CLS] (same as [BOS]) token prepended, and a [SEP] (same as [EOS]) token separating sequences.
35
+ For best results, make sure that [CLS] is prepended to your input sequence (see sample usage linked above)!
36
  Details for this model specifically:
37
 
38
  * Architecture: gpt2
 
58
  author={Chang, Tyler A. and Arnett, Catherine and Tu, Zhuowen and Bergen, Benjamin K.},
59
  journal={Preprint},
60
  year={2024},
61
+ url={https://www.arxiv.org/abs/2408.10441},
62
  }
63
  ```