| name: wordpiece_tokenizer | |
| config_type: preprocessor | |
| pretrained_path: hezar-ai/bert-base-fa | |
| max_length: 512 | |
| truncation_strategy: longest_first | |
| truncation_direction: right | |
| stride: 0 | |
| padding_strategy: longest | |
| padding_direction: right | |
| pad_to_multiple_of: 0 | |
| pad_token_id: 0 | |
| pad_token: '[PAD]' | |
| pad_token_type_id: 0 | |
| unk_token: '[UNK]' | |
| special_tokens: | |
| - '[UNK]' | |
| - '[SEP]' | |
| - '[CLS]' | |
| - '[PAD]' | |
| - '[MASK]' | |
| wordpieces_prefix: '##' | |
| train_config: | |
| name: wordpiece_tokenizer | |
| config_type: preprocessor | |
| vocab_size: 30000 | |
| min_frequency: 2 | |
| limit_alphabet: 1000 | |
| initial_alphabet: [] | |
| show_progress: true | |