From 0fe7add363a34cb73e305ea28fe0823d772b9eb6 Mon Sep 17 00:00:00 2001 From: Taku Kudo Date: Sun, 28 Jan 2024 05:19:00 +0000 Subject: [PATCH] fixed crash bug in unigram model training --- src/unigram_model_trainer.cc | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/unigram_model_trainer.cc b/src/unigram_model_trainer.cc index d921c06f..0f4a580f 100644 --- a/src/unigram_model_trainer.cc +++ b/src/unigram_model_trainer.cc @@ -229,7 +229,7 @@ TrainerModel::SentencePieces Trainer::MakeSeedSentencePiecesInternal() { for (node_int_type i = 0; i < node_num; ++i) { const node_int_type offset = SA[L[i]]; const node_int_type len = D[i]; - if (len <= 1) { + if (len <= 1 || offset >= array.size() || offset + len >= array.size()) { continue; } const char32 *begin = &array[offset];