Welcome to mirror list, hosted at ThFree Co, Russian Federation.

github.com/marian-nmt/sentencepiece.git - Unnamed repository; edit this file 'description' to name the repository.
summaryrefslogtreecommitdiff
path: root/src
diff options
context:
space:
mode:
authorTaku Kudo <taku@google.com>2020-04-24 20:30:13 +0300
committerTaku Kudo <taku@google.com>2020-04-24 20:30:13 +0300
commitb8101c5017cffad26d5c2352a73c9c731aa506a3 (patch)
tree05555e8342c291e3096c2bb48fd2c907b61edb28 /src
parente8a84a16d13e8bf92892a1cd92e4de3b0d0321fd (diff)
Prefer longest user_defined_symbol if ambigous
Diffstat (limited to 'src')
-rw-r--r--src/unigram_model.cc2
1 files changed, 1 insertions, 1 deletions
diff --git a/src/unigram_model.cc b/src/unigram_model.cc
index 083097d..4af85ef 100644
--- a/src/unigram_model.cc
+++ b/src/unigram_model.cc
@@ -403,7 +403,7 @@ void Model::PopulateNodes(Lattice *lattice) const {
Lattice::Node *node = lattice->Insert(begin_pos, length);
node->id = id; // the value of Trie stores vocab_id.
// User defined symbol receives extra bonus to always be selected.
- node->score = IsUserDefinedInlined(id) ? (length * max_score_ + 1.0)
+ node->score = IsUserDefinedInlined(id) ? (length * max_score_ - 0.1)
: GetScoreInlined(id);
if (!has_single_node && node->length == 1) {
has_single_node = true;