precious3-gpt / tokenizer_config.json
stefan-insilico's picture
Loaded model configuration and mpt-7b code
f53ff72
raw
history blame
242 Bytes
{
"bos_token": "[BOS]",
"clean_up_tokenization_spaces": true,
"eos_token": "[EOS]",
"model_max_length": 1000000000000000019884624838656,
"pad_token": "[PAD]",
"tokenizer_class": "PreTrainedTokenizerFast",
"unk_token": "[UNK]"
}