metadata

base_model: bert-base-chinese
tags:
  - generated_from_trainer
model-index:
  - name: bert-base-cantonese
    results: []
license: cc-by-4.0
language:
  - yue
pipeline_tag: fill-mask
widget:
  - text: 香港原本[MASK]一個人煙稀少嘅漁港。
    example_title: 係

bert-base-cantonese

This model is a continue pre-train version of bert-base-chinese on Cantonese Common Crawl dataset with 198m tokens.

Model description

This model has extended 500 more Chinese characters which very common in Cantonese, such as 冧, 噉, 麪, 笪, 冚, 乸 etc.

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 0.0001
train_batch_size: 24
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 192
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 1.0

Training results

Framework versions

Transformers 4.35.0.dev0
Pytorch 2.1.1+cu121
Datasets 2.14.6
Tokenizers 0.14.1