File size: 486 Bytes
87c85b5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
---
language: dv
---

# byt5-base-dv

Pretrained from scratch on Dhivei (language of the Maldives)
with ByT5, Google's new byte-level tokenizer strategy.

Corpus: Sofwath's Dhivehi corpus https://github.com/Sofwath/DhivehiDatasets

Pretraining Notebook: 
https://colab.research.google.com/drive/1ERIZ1PyHn-yN_jo7dTQeODn22vrt-d1d?usp=sharing

## Fine-tuning Demo

On Dhivehi news classification task

https://colab.research.google.com/drive/11u5SafR4bKICmArgDl6KQ9vqfYtDpyWp?usp=sharing