metadata

base_model: /notebooks/cantonese/bert-base-cantonese
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 indiejoseph/wikipedia-zh-yue-filtered.

Model description

This model has extended 500 more Chinese characters which very common in Cantonese, such as 冧, 噉, 麪, 笪, 冚, 乸 etc, and continue pre-trained with indiejoseph/wikipedia-zh-yue-filtered

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: 5e-05
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
num_epochs: 10.0

Training results

Framework versions

Transformers 4.34.0.dev0
Pytorch 2.0.1+cu117
Datasets 2.14.5
Tokenizers 0.14.0