Spaces:
Running
Running
File size: 1,942 Bytes
751936e |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 |
"""
最简单的tokenizer
"""
import json
from tokenizers import Tokenizer
tokenizer = Tokenizer.from_file("20B_tokenizer.json")
print("vocab_size with added_tokens:", tokenizer.get_vocab_size(with_added_tokens=True))
print("vocab_size without added_tokens:", tokenizer.get_vocab_size(with_added_tokens=False))
vocab = tokenizer.get_vocab()
def test_single_token():
"""
单个字符的编码（一个字符可能会编码成多个id）
"""
for word in "发大厦三分赛中国解决方法黑白侗鸩，。！？；":
encoding = tokenizer.encode(word)
for token_id in encoding.ids:
decode_str = tokenizer.decode([token_id]) # 特殊字符解码后会统一变成 �，对应 "\ufffd"
token = tokenizer.id_to_token(token_id)
print(word, token_id, decode_str, json.dumps(decode_str), token, json.dumps(token))
def test_long_token():
"""
"""
words = [
"//----------------------------------------------------------------", # 代码里有
"--------------------------",
"-------------------------",
"-----------------------",
]
for word in words:
encoding = tokenizer.encode(word)
for token_id in encoding.ids:
decode_str = tokenizer.decode([token_id]) #
token = tokenizer.id_to_token(token_id)
print(word, token_id, decode_str, json.dumps(decode_str), token, json.dumps(token))
def test_encode():
text = "中国解决方法黑白侗鸩，。！？；一个人去哪里 一 个"
encoding = tokenizer.encode(text)
for token_id in encoding.ids:
decode_str = tokenizer.decode([token_id]) # 特殊字符解码后会统一变成 �，对应 "\ufffd"
token = tokenizer.id_to_token(token_id)
print(token_id, decode_str, json.dumps(decode_str), token, json.dumps(token))
test_single_token()
# test_long_token()
# test_encode()
|