-
Notifications
You must be signed in to change notification settings - Fork 643
/
tokenizer.py
266 lines (215 loc) · 9.21 KB
/
tokenizer.py
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
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
# take from https://github.com/openai/CLIP/blob/main/clip/simple_tokenizer.py
# to give users a quick easy start to training DALL-E without doing BPE
import torch
import youtokentome as yttm
from tokenizers import Tokenizer
from tokenizers.processors import ByteLevel
from transformers import BertTokenizer
import html
import os
from functools import lru_cache
from pathlib import Path
import ftfy
import regex as re
# OpenAI simple tokenizer
@lru_cache()
def default_bpe():
return os.path.join(os.path.dirname(os.path.abspath(__file__)), "data/bpe_simple_vocab_16e6.txt")
@lru_cache()
def bytes_to_unicode():
bs = list(range(ord("!"), ord("~") + 1)) + list(range(ord("¡"), ord("¬") + 1)) + list(range(ord("®"), ord("ÿ") + 1))
cs = bs[:]
n = 0
for b in range(2 ** 8):
if b not in bs:
bs.append(b)
cs.append(2 ** 8 + n)
n += 1
cs = [chr(n) for n in cs]
return dict(zip(bs, cs))
def get_pairs(word):
pairs = set()
prev_char = word[0]
for char in word[1:]:
pairs.add((prev_char, char))
prev_char = char
return pairs
def basic_clean(text):
text = ftfy.fix_text(text)
text = html.unescape(html.unescape(text))
return text.strip()
def whitespace_clean(text):
text = re.sub(r'\s+', ' ', text)
text = text.strip()
return text
class SimpleTokenizer(object):
def __init__(self, bpe_path = default_bpe()):
self.byte_encoder = bytes_to_unicode()
self.byte_decoder = {v: k for k, v in self.byte_encoder.items()}
merges = Path(bpe_path).read_text(encoding='utf8').split('\n')
merges = merges[1:49152 - 256 - 2 + 1]
merges = [tuple(merge.split()) for merge in merges]
vocab = list(bytes_to_unicode().values())
vocab = vocab + [v + '</w>' for v in vocab]
for merge in merges:
vocab.append(''.join(merge))
vocab.extend(['<|startoftext|>', '<|endoftext|>'])
self.vocab_size = 49408
self.encoder = dict(zip(vocab, range(len(vocab))))
self.decoder = {v: k for k, v in self.encoder.items()}
self.bpe_ranks = dict(zip(merges, range(len(merges))))
self.cache = {'<|startoftext|>': '<|startoftext|>', '<|endoftext|>': '<|endoftext|>'}
self.pat = re.compile(
r"""<\|startoftext\|>|<\|endoftext\|>|'s|'t|'re|'ve|'m|'ll|'d|[\p{L}]+|[\p{N}]|[^\s\p{L}\p{N}]+""",
re.IGNORECASE)
def bpe(self, token):
if token in self.cache:
return self.cache[token]
word = tuple(token[:-1]) + (token[-1] + '</w>',)
pairs = get_pairs(word)
if not pairs:
return token + '</w>'
while True:
bigram = min(pairs, key=lambda pair: self.bpe_ranks.get(pair, float('inf')))
if bigram not in self.bpe_ranks:
break
first, second = bigram
new_word = []
i = 0
while i < len(word):
try:
j = word.index(first, i)
new_word.extend(word[i:j])
i = j
except:
new_word.extend(word[i:])
break
if word[i] == first and i < len(word) - 1 and word[i + 1] == second:
new_word.append(first + second)
i += 2
else:
new_word.append(word[i])
i += 1
new_word = tuple(new_word)
word = new_word
if len(word) == 1:
break
else:
pairs = get_pairs(word)
word = ' '.join(word)
self.cache[token] = word
return word
def encode(self, text):
bpe_tokens = []
text = whitespace_clean(basic_clean(text)).lower()
for token in re.findall(self.pat, text):
token = ''.join(self.byte_encoder[b] for b in token.encode('utf-8'))
bpe_tokens.extend(self.encoder[bpe_token] for bpe_token in self.bpe(token).split(' '))
return bpe_tokens
def decode(self, tokens, remove_start_end = True, pad_tokens = set()):
if torch.is_tensor(tokens):
tokens = tokens.tolist()
if remove_start_end:
tokens = [token for token in tokens if token not in (49406, 40407, 0)]
text = ''.join([self.decoder[token] for token in tokens if token not in pad_tokens])
text = bytearray([self.byte_decoder[c] for c in text]).decode('utf-8', errors="replace").replace('</w>', ' ')
return text
def tokenize(self, texts, context_length = 256, truncate_text = False):
if isinstance(texts, str):
texts = [texts]
all_tokens = [self.encode(text) for text in texts]
result = torch.zeros(len(all_tokens), context_length, dtype=torch.long)
for i, tokens in enumerate(all_tokens):
if len(tokens) > context_length:
if truncate_text:
tokens = tokens[:context_length]
else:
raise RuntimeError(f"Input {texts[i]} is too long for context length {context_length}")
result[i, :len(tokens)] = torch.tensor(tokens)
return result
tokenizer = SimpleTokenizer()
# huggingface tokenizer
class HugTokenizer:
def __init__(self, bpe_path = None):
bpe_path = Path(bpe_path)
assert bpe_path.exists(), f'BPE json path {str(bpe_path)} does not exist'
tokenizer = Tokenizer.from_file(str(bpe_path))
tokenizer.post_processor = ByteLevel(trim_offsets = True)
self.tokenizer = tokenizer
self.vocab_size = tokenizer.get_vocab_size()
def decode(self, tokens, pad_tokens = set()):
if torch.is_tensor(tokens):
tokens = tokens.tolist()
ignore_ids = pad_tokens.union({0})
tokens = [token for token in tokens if token not in ignore_ids]
return self.tokenizer.decode(tokens, skip_special_tokens = True)
def encode(self, text):
return self.tokenizer.encode(text).ids
def tokenize(self, texts, context_length = 256, truncate_text = False):
if isinstance(texts, str):
texts = [texts]
all_tokens = [self.encode(text) for text in texts]
result = torch.zeros(len(all_tokens), context_length, dtype=torch.long)
for i, tokens in enumerate(all_tokens):
if len(tokens) > context_length:
if truncate_text:
tokens = tokens[:context_length]
else:
raise RuntimeError(f"Input {texts[i]} is too long for context length {context_length}")
result[i, :len(tokens)] = torch.tensor(tokens)
return result
# chinese tokenizer
class ChineseTokenizer:
def __init__(self):
tokenizer = BertTokenizer.from_pretrained('bert-base-chinese')
self.tokenizer = tokenizer
self.vocab_size = tokenizer.vocab_size
def decode(self, tokens, pad_tokens = set()):
if torch.is_tensor(tokens):
tokens = tokens.tolist()
ignore_ids = pad_tokens.union({0})
tokens = [token for token in tokens if token not in ignore_ids]
return self.tokenizer.decode(tokens)
def encode(self, text):
return torch.tensor(self.tokenizer.encode(text, add_special_tokens = False))
def tokenize(self, texts, context_length = 256, truncate_text = False):
if isinstance(texts, str):
texts = [texts]
all_tokens = [self.encode(text) for text in texts]
result = torch.zeros(len(all_tokens), context_length, dtype=torch.long)
for i, tokens in enumerate(all_tokens):
if len(tokens) > context_length:
if truncate_text:
tokens = tokens[:context_length]
else:
raise RuntimeError(f"Input {texts[i]} is too long for context length {context_length}")
result[i, :len(tokens)] = torch.tensor(tokens)
return result
# yttm tokenizer
class YttmTokenizer:
def __init__(self, bpe_path = None):
bpe_path = Path(bpe_path)
assert bpe_path.exists(), f'BPE json path {str(bpe_path)} does not exist'
tokenizer = yttm.BPE(model = str(bpe_path))
self.tokenizer = tokenizer
self.vocab_size = tokenizer.vocab_size()
def decode(self, tokens, pad_tokens = set()):
if torch.is_tensor(tokens):
tokens = tokens.tolist()
return self.tokenizer.decode(tokens, ignore_ids = pad_tokens.union({0}))
def encode(self, texts):
encoded = self.tokenizer.encode(texts, output_type = yttm.OutputType.ID)
return list(map(torch.tensor, encoded))
def tokenize(self, texts, context_length = 256, truncate_text = False):
if isinstance(texts, str):
texts = [texts]
all_tokens = self.encode(texts)
result = torch.zeros(len(all_tokens), context_length, dtype=torch.long)
for i, tokens in enumerate(all_tokens):
if len(tokens) > context_length:
if truncate_text:
tokens = tokens[:context_length]
else:
raise RuntimeError(f"Input {texts[i]} is too long for context length {context_length}")
result[i, :len(tokens)] = torch.tensor(tokens)
return result