lsg-albert-base-v2-4096 / tokenizer_config.json
ccdv's picture
first
2d1da30
{
"bos_token": "[CLS]",
"cls_token": "[CLS]",
"do_lower_case": true,
"eos_token": "[SEP]",
"keep_accents": false,
"mask_token": {
"__type": "AddedToken",
"content": "[MASK]",
"lstrip": true,
"normalized": false,
"rstrip": false,
"single_word": false
},
"model_max_length": 4096,
"name_or_path": "albert-base-v2",
"pad_token": "<pad>",
"remove_space": true,
"sep_token": "[SEP]",
"special_tokens_map_file": null,
"tokenizer_class": "AlbertTokenizer",
"unk_token": "<unk>"
}