parakeet-ctc-1.1b / tokenizer_config.json
nithinraok's picture
trfms-integration (#2)
a707e81 verified
raw
history blame contribute delete
634 Bytes
{
"added_tokens_decoder": {
"0": {
"content": "<unk>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"1024": {
"content": "<pad>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
}
},
"clean_up_tokenization_spaces": false,
"extra_special_tokens": {},
"model_max_length": 1000000000000000019884624838656,
"pad_token": "<pad>",
"processor_class": "ParakeetProcessor",
"tokenizer_class": "ParakeetTokenizerFast",
"unk_token": "<unk>"
}