common-voice-lithuanian-fairseq / tokenizer_config.json
birgermoell's picture
Create tokenizer_config.json
7adc27d
raw
history blame contribute delete
304 Bytes
{"unk_token": "[UNK]", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "[PAD]", "do_lower_case": false, "word_delimiter_token": "|", "replace_word_delimiter_char": " ", "special_tokens_map_file": null, "name_or_path": "./wav2vec2-common_voice-lithuanian", "tokenizer_class": "Wav2Vec2CTCTokenizer"}