Support getting word IDs for CTC HLG decoding. (#978)

2024-06-06 14:22:39 +08:00
parent 69347ffc8f
commit 1a43d1e37f
13 changed files with 60 additions and 13 deletions
--- a/sherpa-onnx/csrc/offline-stream.h
+++ b/sherpa-onnx/csrc/offline-stream.h
@@ -30,6 +30,8 @@ struct OfflineRecognitionResult {
  /// timestamps[i] records the time in seconds when tokens[i] is decoded.
  std::vector<float> timestamps;

+  std::vector<int32_t> words;
+
  std::string AsJsonString() const;
 };