Skip to content

GitLab

Commit a80cade9 authored by Myle Ott's avatar Myle Ott Committed by Facebook Github Bot
Browse files

Update BPE library code 

Summary: Pull Request resolved: https://github.com/fairinternal/fairseq-py/pull/780

Differential Revision: D16537567

Pulled By: myleott

fbshipit-source-id: 4e18c529959935e82ea122c3a2ee477308ffcbe3
parent 76ff39f5
Show whitespace changes
Inline Side-by-side
Showing with 132 additions and 124 deletions
fairseq/data/encoders/gpt2_bpe.py
View file @ a80cade9
......@@ -8,6 +8,8 @@
from fairseq import file_utils
from fairseq.data.encoders import register_bpe
from .gpt2_bpe_utils import get_encoder
DEFAULT_ENCODER_JSON = 'https://dl.fbaipublicfiles.com/fairseq/gpt2_bpe/encoder.json'
DEFAULT_VOCAB_BPE = 'https://dl.fbaipublicfiles.com/fairseq/gpt2_bpe/vocab.bpe'
......@@ -42,127 +44,5 @@ class GPT2BPE(object):
def decode(self, x: str) -> str:
return self.bpe.decode(map(int, x.split()))
"""Byte pair encoding utilities from GPT-2"""
from functools import lru_cache
import json
import os
@lru_cache()
def bytes_to_unicode():
"""
Returns list of utf-8 byte and a corresponding list of unicode strings.
The reversible bpe codes work on unicode strings.
This means you need a large # of unicode characters in your vocab if you want to avoid UNKs.
When you're at something like a 10B token dataset you end up needing around 5K for decent coverage.
This is a signficant percentage of your normal, say, 32K bpe vocab.
To avoid that, we want lookup tables between utf-8 bytes and unicode strings.
And avoids mapping to whitespace/control characters the bpe code barfs on.
"""
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):
"""Return set of symbol pairs in a word.
Word is represented as tuple of symbols (symbols being variable-length strings).
"""
pairs = set()
prev_char = word[0]
for char in word[1:]:
pairs.add((prev_char, char))
prev_char = char
return pairs
class Encoder:
def __init__(self, encoder, bpe_merges, errors='replace'):
self.encoder = encoder
self.decoder = {v:k for k,v in self.encoder.items()}
self.errors = errors # how to handle errors in decoding
self.byte_encoder = bytes_to_unicode()
self.byte_decoder = {v:k for k, v in self.byte_encoder.items()}
self.bpe_ranks = dict(zip(bpe_merges, range(len(bpe_merges))))
self.cache = {}
try:
import regex as re
self.re = re
except ImportError:
raise ImportError('Please install regex with: pip install regex')
# Should haved added re.IGNORECASE so BPE merges can happen for capitalized versions of contractions
self.pat = self.re.compile(r"""'s|'t|'re|'ve|'m|'ll|'d| ?\p{L}+| ?\p{N}+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+""")
def bpe(self, token):
if token in self.cache:
return self.cache[token]
word = tuple(token)
pairs = get_pairs(word)
if not pairs:
return token
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 = []
for token in self.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):
text = ''.join([self.decoder[token] for token in tokens])
text = bytearray([self.byte_decoder[c] for c in text]).decode('utf-8', errors=self.errors)
return text
def get_encoder(encoder_json_path, vocab_bpe_path):
with open(encoder_json_path, 'r') as f:
encoder = json.load(f)
with open(vocab_bpe_path, 'r', encoding="utf-8") as f:
bpe_data = f.read()
bpe_merges = [tuple(merge_str.split()) for merge_str in bpe_data.split('\n')[1:-1]]
return Encoder(
encoder=encoder,
bpe_merges=bpe_merges,
)
def is_beginning_of_word(self, x: str) -> bool:
return self.decode(x).startswith(' ')
fairseq/data/encoders/gpt2_bpe_utils.py 0 → 100644
View file @ a80cade9
"""
Byte pair encoding utilities from GPT-2.
Original source: https://github.com/openai/gpt-2/blob/master/src/encoder.py
Original license: MIT
"""
from functools import lru_cache
import json
import os
@lru_cache()
def bytes_to_unicode():
"""
Returns list of utf-8 byte and a corresponding list of unicode strings.
The reversible bpe codes work on unicode strings.
This means you need a large # of unicode characters in your vocab if you want to avoid UNKs.
When you're at something like a 10B token dataset you end up needing around 5K for decent coverage.
This is a signficant percentage of your normal, say, 32K bpe vocab.
To avoid that, we want lookup tables between utf-8 bytes and unicode strings.
And avoids mapping to whitespace/control characters the bpe code barfs on.
"""
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):
"""Return set of symbol pairs in a word.
Word is represented as tuple of symbols (symbols being variable-length strings).
"""
pairs = set()
prev_char = word[0]
for char in word[1:]:
pairs.add((prev_char, char))
prev_char = char
return pairs
class Encoder:
def __init__(self, encoder, bpe_merges, errors='replace'):
self.encoder = encoder
self.decoder = {v:k for k,v in self.encoder.items()}
self.errors = errors # how to handle errors in decoding
self.byte_encoder = bytes_to_unicode()
self.byte_decoder = {v:k for k, v in self.byte_encoder.items()}
self.bpe_ranks = dict(zip(bpe_merges, range(len(bpe_merges))))
self.cache = {}
try:
import regex as re
self.re = re
except ImportError:
raise ImportError('Please install regex with: pip install regex')
# Should haved added re.IGNORECASE so BPE merges can happen for capitalized versions of contractions
self.pat = self.re.compile(r"""'s|'t|'re|'ve|'m|'ll|'d| ?\p{L}+| ?\p{N}+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+""")
def bpe(self, token):
if token in self.cache:
return self.cache[token]
word = tuple(token)
pairs = get_pairs(word)
if not pairs:
return token
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 = []
for token in self.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):
text = ''.join([self.decoder[token] for token in tokens])
text = bytearray([self.byte_decoder[c] for c in text]).decode('utf-8', errors=self.errors)
return text
def get_encoder(encoder_json_path, vocab_bpe_path):
with open(encoder_json_path, 'r') as f:
encoder = json.load(f)
with open(vocab_bpe_path, 'r', encoding="utf-8") as f:
bpe_data = f.read()
bpe_merges = [tuple(merge_str.split()) for merge_str in bpe_data.split('\n')[1:-1]]
return Encoder(
encoder=encoder,
bpe_merges=bpe_merges,
)
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment