roberta-base-catalan / tokenizer_config.json
ClassCat's picture
add tokenizer
05b4b46
raw
history blame
429 Bytes
{"errors": "replace", "bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "cls_token": "<s>", "unk_token": "<unk>", "pad_token": "<pad>", "mask_token": "<mask>", "add_prefix_space": false, "trim_offsets": true, "keep_accents": true, "max_len": 50, "special_tokens_map_file": "classcat/roberta-base-catalan/special_tokens_map.json", "name_or_path": "classcat/roberta-base-catalan", "tokenizer_class": "RobertaTokenizer"}