vina-cased-tokenizer / tokenizer_config.json
phunc20's picture
Upload tokenizer
270036c
raw
history blame
506 Bytes
{
"add_prefix_space": false,
"bos_token": "<s>",
"cls_token": "<s>",
"eos_token": "</s>",
"errors": "replace",
"mask_token": "<mask>",
"model_max_length": 1000000000000000019884624838656,
"name_or_path": "/home/phunc20/downloads/vina-cased_tokenizer",
"pad_token": "<pad>",
"sep_token": "</s>",
"special_tokens_map_file": "/home/phunc20/downloads/vina-cased_tokenizer/special_tokens_map.json",
"tokenizer_class": "RobertaTokenizer",
"trim_offsets": true,
"unk_token": "<unk>"
}