Commits

History for

Commits on Sep 26, 2026

fix(python): accept out_type in DecodeIds and DecodePieces
kwy404
committed
e07470f

Commits on Sep 25, 2026

fix(python): bind SentencePieceNormalizer.LoadFromSerializedNormalizerSpec
kwy404
committed
de2c7d2

Commits on Sep 15, 2026

feat(lite): introduce SentencePiece Lite zero-dependency C++20 runtime
taku910
committed
a2cb4d8

Commits on Sep 12, 2026

Promote auto_character_coverage to TrainerSpec and remove fixed_sparse_lambda
taku910
committed
df18339

Commits on Sep 10, 2026

feat,fix,deprecate: allow unk_id=-1 with byte_fallback, fix issue #1308, and retire deprecated flags
taku910
committed
c2fbf86

Commits on Sep 8, 2026

Merge pull request #1320 from Kayvan-Zahiri/fix/offset-mapping-empty-batch
taku910
authored
73666a8
Update sentencepiece_test.py
taku910
authored
8d26982

Commits on Sep 3, 2026

Commits on Sep 2, 2026

Fix encode() with an empty batch and out_type='offset_mapping'
Kayvan-Zahiri
committed
63b1965

Commits on Aug 30, 2026

build: upgrade MSVC and common C++ standard from C++17 to C++20
taku910
committed
7520a37

Commits on Aug 26, 2026

cleanup: remove self_test_data and consolidate deprecated fields in proto
taku910
committed
470b7d5

Commits on Aug 14, 2026

remove unused trian method from python module
taku910
committed
8f4966b

Commits on Aug 13, 2026

test: standardize test model file names
taku910
committed
1a1126e
build: clean up build scripts, CI workflow, and gitignore
taku910
committed
df4e7cb

Commits on Aug 12, 2026

build: fix cmake dependencies, sdist self-containment, and wheel CI workflow
taku910
committed
4e5215e

Commits on Aug 10, 2026

build: fix utf8_range include path for protobuf v35+ in cmake and setup.py
taku910
committed
5cc3786
use tempfile.TemporaryDirectory in test
taku910
committed
76d28d6

Commits on Aug 1, 2026

fixed the slow mac wheel build
taku910
committed
fdcc3e5

Commits on Jul 31, 2026

Merge remote-tracking branch 'origin/master' into l1_sparse
taku910
committed
5689abc
fixed build error on windows
taku910
committed
2f4a70f
Merge remote-tracking branch 'origin/master' into l1_sparse
taku910
committed
39b7f65
fixed build error on windows
taku910
committed
8f01700
Merge remote-tracking branch 'origin/master' into l1_sparse
taku910
committed
b3eadd7
fixed build error on windows
taku910
committed
e07dd00

Commits on Jul 29, 2026

Update Python module build for Protobuf v25.6 and C++20
taku910
committed
dc1f89c

Commits on Jul 28, 2026

Refactor Unigram L1 sparse pruning, Post-Lasso debiased refit, and Python bindings
taku910
committed
fd53b5c

Commits on Jul 22, 2026

Remove unused includes from pybind.cc
taku910
committed
28d5b00

Commits on Jul 18, 2026

Refactor Trainer and pretokenization interface
taku910
committed
d1b6696

Commits on Jul 17, 2026

Fix Decode() of empty input for non-str return types
sohumt123
committed
2056873

Commits on Jul 15, 2026

Commits on Jun 22, 2026

Stop support for Win32 wheels
taku910
committed
24fd00c