File size: 515 Bytes
751936e
 
 
 
da93e39
 
 
751936e
da93e39
751936e
 
 
f4973d4
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

import os
from transformers import AutoTokenizer, BloomTokenizerFast

CURRENT_DIR = os.path.dirname(os.path.abspath(__file__))
TOKENIZER_DIR = os.path.join(CURRENT_DIR, "moss-moon-003-sft")
tokenizer = AutoTokenizer.from_pretrained(TOKENIZER_DIR, trust_remote_code=True)

# tokenizer = AutoTokenizer.from_pretrained("fnlp/moss-moon-003-sft", trust_remote_code=True)

# vocab_size = len(tokenizer.get_vocab())
# vocab_size = tokenizer.vocab_size

tokenizer.comments = "在gpt2词典基础上,扩充了5万中文"