Skip to content

Improve BERT tokenization for accented characters and non-latin scripts #5740

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
merged 5 commits into from
Feb 28, 2024
Merged

Improve BERT tokenization for accented characters and non-latin scripts #5740

Changes from 1 commit
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
9c996e3
implement nfd for stripping accents in wpm tokenizer
iamlemec Feb 26, 2024
2dd36d6
sort nfd map; reuse iterator
iamlemec Feb 27, 2024
6b33a09
use builtin tolower
iamlemec Feb 27, 2024
9242cf1
add locale include
iamlemec Feb 27, 2024
801abe5
Simplify to_lower cases
iamlemec Feb 27, 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
17 changes: 9 additions & 8 deletions llama.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -68,6 +68,7 @@
#include <cstdio>
#include <cstring>
#include <ctime>
#include <cwctype>
#include <forward_list>
#include <fstream>
#include <functional>
Expand Down Expand Up @@ -8955,14 +8956,14 @@ struct llm_tokenizer_wpm {
}

uint32_t to_lower(uint32_t code) {
if (
(code >= 0x041 && code <= 0x05A) || // latin
(code >= 0x391 && code <= 0x3A9) || // greek
(code >= 0x410 && code <= 0x42F) // cyrillic
) {
return code + 32;
}
return code;
#if defined(_WIN32)
if (code > 0xFFFF) {
return code;
}
return std::tolower(static_cast<wchar_t>(code), std::locale("en_US.UTF-8"));
#else
return std::tolower((wchar_t)code, std::locale("en_US.UTF-8"));
#endif
}

bool is_ascii_punct(uint32_t code) {
Expand Down