mpt-7b-tokenizer / tokenizer_config.json
main-horse's picture
Update tokenizer_config.json
dbd009d
{
"add_prefix_space": false,
"boi_token": "<info>",
"bos_token": "<|endoftext|>",
"clean_up_tokenization_spaces": true,
"eoi_token": "</info>",
"eos_token": "<|endoftext|>",
"model_max_length": 9999999,
"tokenizer_class": "GPTNeoXTokenizer",
"unk_token": "<|endoftext|>",
"pad_token": "<|padding|>"
}