xlm-roberta-long-base-4096 / tokenizer_config.json
AshtonIsNotHere's picture
Added files to fix tokenizer problems
fd32196
raw
history blame
212 Bytes
{"bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "cls_token": "<s>", "unk_token": "<unk>", "pad_token": "<pad>", "mask_token": "<mask>", "model_max_length": 4096, "name_or_path": "xlm-roberta-base"}