Skip to content

Bump tokenizers submodule to fix sentencepiece GCC 15 build #20135

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
rascani merged 2 commits into from
Jun 9, 2026
+6 −2
Merged

Bump tokenizers submodule to fix sentencepiece GCC 15 build #20135

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
82463ae
Bump tokenizers submodule to fix sentencepiece GCC 15 build
rascani Jun 8, 2026
bc3e7cd
Fix Parakeet tokenizer C++20 build
rascani Jun 9, 2026
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
6 changes: 5 additions & 1 deletion examples/models/parakeet/tokenizer_utils.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,10 @@

namespace {

// SentencePiece's word-boundary marker, spelled as UTF-8 bytes so this remains
// a const char[] literal when compiled as C++20.
constexpr char kSentencePieceWordBoundary[] = "\xE2\x96\x81";

bool is_whitespace_only(const std::string& token) {
if (token.empty()) {
return true;
Expand Down Expand Up @@ -36,7 +40,7 @@ bool is_special_token(const std::string& token) {
if (token.rfind("##", 0) == 0) {
return true;
}
if (token.rfind(u8"▁", 0) == 0) {
if (token.rfind(kSentencePieceWordBoundary, 0) == 0) {
return true;
}
if (is_whitespace_only(token)) {
Expand Down
2 changes: 1 addition & 1 deletion extension/llm/tokenizers
Open in desktop
Submodule tokenizers updated 28 files
+30 −0 .ci/scripts/wheel/pre_build_script.sh
+85 −0 .ci/scripts/wheel/test_wheel.py
+39 −0 .ci/scripts/wheel/vc_env_helper.bat
+51 −0 .github/workflows/build-wheels-linux-aarch64.yml
+50 −0 .github/workflows/build-wheels-linux.yml
+52 −0 .github/workflows/build-wheels-macos-arm64.yml
+51 −0 .github/workflows/build-wheels-windows.yml
+22 −0 .github/workflows/claude-code.yml
+1 −1 CMakeLists.txt
+0 −1 include/pytorch/tokenizers/pcre2_regex.h
+41 −44 include/pytorch/tokenizers/string_integer_map.h
+1 −1 setup.py
+64 −20 src/bpe_tokenizer_base.cpp
+19 −13 src/hf_tokenizer.cpp
+42 −11 src/llama2c_tokenizer.cpp
+2 −1 src/normalizer.cpp
+15 −15 src/pcre2_regex.cpp
+13 −10 src/post_processor.cpp
+0 −1 src/regex_lookahead.cpp
+0 −2 src/tekken.cpp
+2 −6 src/tiktoken.cpp
+1 −1 test/CMakeLists.txt
+15 −0 test/targets.bzl
+152 −0 test/test_concurrent_encode.cpp
+124 −0 test/test_llama2c_tokenizer.cpp
+0 −1 test/test_regex.cpp
+13 −0 test/test_tiktoken.cpp
+1 −1 third-party/sentencepiece
Loading