legal-xlm-roberta-base / tokenizer_config.json
joelniklaus's picture
Training in progress, step 50000
498d0ad
raw
history blame
390 Bytes
{
"bos_token": "<s>",
"cls_token": "<s>",
"eos_token": "</s>",
"mask_token": "<mask>",
"model_max_length": 512,
"name_or_path": "data/plms/legal-xlm-roberta-base",
"pad_token": "<pad>",
"sep_token": "</s>",
"special_tokens_map_file": "data/plms/legal-xlm-roberta-base_128k/special_tokens_map.json",
"tokenizer_class": "PreTrainedTokenizerFast",
"unk_token": "<unk>"
}