mpt-7b-wizardlm / tokenizer_config.json
winglian's picture
add base configuration and py code from base model
a583ee9
raw
history blame
237 Bytes
{
"add_prefix_space": false,
"bos_token": "<|endoftext|>",
"clean_up_tokenization_spaces": true,
"eos_token": "<|endoftext|>",
"model_max_length": 2048,
"tokenizer_class": "GPTNeoXTokenizer",
"unk_token": "<|endoftext|>"
}