SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

BR_BERTo model card (#6793)

Browse Source
This commit is contained in:
Rodolfo De Nadai
2020-08-30 08:02:46 -03:00
committed by GitHub
parent d176aaad7f
commit 0eecaceac7
1 changed files with 9 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
model_cards/rdenadai/BR_BERTo/README.md
View File

@@ -14,13 +14,17 @@ Portuguese (Brazil) model for text inference.
## Params ## Params
Trained on a corpus of 5_258_624 sentences, with 132_807_374 non unique tokens (992_418 unique tokens). Trained on a corpus of 6_993_330 sentences.
- Vocab size: 220_000 - Vocab size: 150_000
- RobertaForMaskedLM size : 32 - RobertaForMaskedLM size : 512
- Num train epochs: 2 - Num train epochs: 3
- Time to train: ~23hs (on GCP with a Nvidia T4) - Time to train: ~10days (on GCP with a Nvidia T4)
I follow the great tutorial from HuggingFace team: I follow the great tutorial from HuggingFace team:
[How to train a new language model from scratch using Transformers and Tokenizers](https://huggingface.co/blog/how-to-train) [How to train a new language model from scratch using Transformers and Tokenizers](https://huggingface.co/blog/how-to-train)
More infor here:
[BR_BERTo](https://github.com/rdenadai/BR-BERTo)