Files
omniASR-CTC-1B-v2/tokenizer_config.json

10 lines
233 B
JSON
Raw Permalink Normal View History

{
"tokenizer_class": "Wav2Vec2CTCTokenizer",
"unk_token": "<unk>",
"bos_token": "<s>",
"eos_token": "</s>",
"pad_token": "<s>",
"word_delimiter_token": " ",
"do_lower_case": false,
"replace_word_delimiter_char": " "
}