File size: 1,037 Bytes
4830e28 2b6cbe6 4830e28 00d9f1a 4830e28 00d9f1a 2b6cbe6 00d9f1a 4830e28 00d9f1a 4830e28 00d9f1a 4830e28 2b6cbe6 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 |
# dv-wave
This is a second attempt at a Dhivehi language model trained with
Google Research's [ELECTRA](https://github.com/google-research/electra).
Tokenization and pre-training CoLab: https://colab.research.google.com/drive/1ZJ3tU9MwyWj6UtQ-8G7QJKTn-hG1uQ9v?usp=sharing
Using SimpleTransformers to classify news https://colab.research.google.com/drive/1KnyQxRNWG_yVwms_x9MUAqFQVeMecTV7?usp=sharing
V1: similar performance to mBERT on news classification task after finetuning for 3 epochs (52%)
V2: fixed tokenizers do_lower_case=False and strip_accents=False to preserve vowel signs of Dhivehi
dv-wave: 89% to mBERT: 52%
## Corpus
Trained on @Sofwath's 307MB corpus of Dhivehi text: https://github.com/Sofwath/DhivehiDatasets
This repo also contains the news classification task CSV
[OSCAR](https://oscar-corpus.com/) was considered but has not been added to pretraining; as of
this writing their web crawl has 126MB of Dhivehi text (79MB deduped).
## Vocabulary
Included as vocab.txt in the upload - vocab_size is 29874
|