Qwen-14B-Chat-LLaMAfied / tokenizer_config.json
hiyouga's picture
update tokenizer
1f90317
raw
history blame
286 Bytes
{
"add_prefix_space": false,
"add_bos_token": false,
"tokenizer_class": "GPT2Tokenizer",
"unk_token": "<|endoftext|>",
"bos_token": "<|endoftext|>",
"eos_token": "<|endoftext|>",
"model_max_length": 1000000000000000019884624838656,
"clean_up_tokenization_spaces": true
}