layazh-zh / tokenizer /tokenizer_config.json
NovaAI
layazh-zh: Chinese System-1 decision model (Laya recipe, RoBERTa-large encoder + decision head); full-eval acc 0.6951 vs upstream 0.4783
f69ec12 verified
Raw History Blame Contribute Delete
394 Bytes
{
"backend": "tokenizers",
"cls_token": "[CLS]",
"do_lower_case": true,
"is_local": false,
"local_files_only": false,
"mask_token": "[MASK]",
"model_max_length": 1000000000000000019884624838656,
"pad_token": "[PAD]",
"sep_token": "[SEP]",
"strip_accents": null,
"tokenize_chinese_chars": true,
"tokenizer_class": "BertTokenizer",
"unk_token": "[UNK]"
}