korean-gpt-quick-test / tokenizer_config.json
prismdata's picture
Add tokenizer and update model
34811aa verified
raw
history blame contribute delete
278 Bytes
{
"tokenizer_class": "KoreanGPTTokenizer",
"auto_map": {
"AutoTokenizer": [
"tokenization_korean_gpt.KoreanGPTTokenizer",
null
]
},
"model_max_length": 512,
"bos_token": "<s>",
"eos_token": "</s>",
"unk_token": "<unk>",
"pad_token": "<pad>"
}