Skip to content

1) Optimize regular expressions used for splitting by ~20% #234

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
hauntsaninja merged 3 commits into from
Feb 9, 2024
Merged

1) Optimize regular expressions used for splitting by ~20% #234

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
9d4d220
Optimize regular expressions used for splitting
Dec 31, 2023
2f04faa
Merge branch 'main' into paplorinc/optimize-regex
l0rinc Jan 30, 2024
6f261de
gpt-2 docs
hauntsaninja Feb 9, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
13 changes: 8 additions & 5 deletions tiktoken_ext/openai_public.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,7 +17,10 @@ def gpt2():
return {
"name": "gpt2",
"explicit_n_vocab": 50257,
"pat_str": r"""'s|'t|'re|'ve|'m|'ll|'d| ?\p{L}+| ?\p{N}+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+""",
# The pattern in the original GPT-2 release is:
# r"""'s|'t|'re|'ve|'m|'ll|'d| ?[\p{L}]+| ?[\p{N}]+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+"""
# This is equivalent, but executes faster:
"pat_str": r"""'(?:[sdmt]|ll|ve|re)| ?\p{L}+| ?\p{N}+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+""",
"mergeable_ranks": mergeable_ranks,
"special_tokens": {ENDOFTEXT: 50256},
}
Expand All @@ -31,7 +34,7 @@ def r50k_base():
return {
"name": "r50k_base",
"explicit_n_vocab": 50257,
"pat_str": r"""'s|'t|'re|'ve|'m|'ll|'d| ?\p{L}+| ?\p{N}+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+""",
"pat_str": r"""'(?:[sdmt]|ll|ve|re)| ?\p{L}+| ?\p{N}+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+""",
"mergeable_ranks": mergeable_ranks,
"special_tokens": {ENDOFTEXT: 50256},
}
Expand All @@ -45,7 +48,7 @@ def p50k_base():
return {
"name": "p50k_base",
"explicit_n_vocab": 50281,
"pat_str": r"""'s|'t|'re|'ve|'m|'ll|'d| ?\p{L}+| ?\p{N}+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+""",
"pat_str": r"""'(?:[sdmt]|ll|ve|re)| ?\p{L}+| ?\p{N}+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+""",
"mergeable_ranks": mergeable_ranks,
"special_tokens": {ENDOFTEXT: 50256},
}
Expand All @@ -59,7 +62,7 @@ def p50k_edit():
special_tokens = {ENDOFTEXT: 50256, FIM_PREFIX: 50281, FIM_MIDDLE: 50282, FIM_SUFFIX: 50283}
return {
"name": "p50k_edit",
"pat_str": r"""'s|'t|'re|'ve|'m|'ll|'d| ?\p{L}+| ?\p{N}+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+""",
"pat_str": r"""'(?:[sdmt]|ll|ve|re)| ?\p{L}+| ?\p{N}+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+""",
"mergeable_ranks": mergeable_ranks,
"special_tokens": special_tokens,
}
Expand All @@ -79,7 +82,7 @@ def cl100k_base():
}
return {
"name": "cl100k_base",
"pat_str": r"""(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\r\n\p{L}\p{N}]?\p{L}+|\p{N}{1,3}| ?[^\s\p{L}\p{N}]+[\r\n]*|\s*[\r\n]+|\s+(?!\S)|\s+""",
"pat_str": r"""'(?i:[sdmt]|ll|ve|re)|[^\r\n\p{L}\p{N}]?+\p{L}+|\p{N}{1,3}| ?[^\s\p{L}\p{N}]++[\r\n]*|\s*[\r\n]|\s+(?!\S)|\s+""",
Copy link
Contributor Author

@l0rinc l0rinc Jan 7, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Besides the contractions that were grouped in the other regexes, here I've optimized using possessive quantifiers to avoid backtracking.
The changes were guided by JMH benchmarks for the same regex: https://github.com/knuddelsgmbh/jtokkit/pull/75/files#r1434984132

"mergeable_ranks": mergeable_ranks,
"special_tokens": special_tokens,
}
Expand Down