PlasmidLM / tokenizer_config.json
McClain's picture
Add tokenizer_config.json for AutoTokenizer compatibility
23f2016 verified
raw
history blame contribute delete
328 Bytes
{
"auto_map": {
"AutoTokenizer": [
"tokenization_plasmid_lm.PlasmidLMTokenizer",
null
]
},
"tokenizer_class": "PlasmidLMTokenizer",
"bos_token": "<BOS>",
"eos_token": "<EOS>",
"unk_token": "<UNK>",
"pad_token": "<PAD>",
"sep_token": "<SEP>",
"model_max_length": 16384,
"use_fast": false
}