LongHeRo / tokenizer_config.json
vitvit's picture
Upload 7 files
a5156f0
raw
history blame
576 Bytes
{
"add_prefix_space": false,
"bos_token": "<s>",
"cls_token": "<s>",
"eos_token": "</s>",
"errors": "replace",
"mask_token": {
"__type": "AddedToken",
"content": "<mask>",
"lstrip": true,
"normalized": false,
"rstrip": false,
"single_word": false
},
"model_max_length": 4096,
"name_or_path": "/home/vitalysh/Projects/HeLM/resources/Longformer/longformer-he",
"pad_token": "<pad>",
"sep_token": "</s>",
"special_tokens_map_file": null,
"tokenizer_class": "LongformerTokenizer",
"trim_offsets": true,
"unk_token": "<unk>"
}