EngineX-Iluvatar/enginex-mr_series-sherpa-onnx
Archived
16
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

Code refactoring (#74)

Browse Source 
* Don't reset model state and feature extractor on endpointing

* support passing decoding_method from commandline

* Add modified_beam_search to Python API

* fix C API example

* Fix style issues
This commit is contained in:
Fangjun Kuang
2023-03-03 12:10:59 +08:00
committed by GitHub
parent c241f93c40
commit 7f72c13d9a
34 changed files with 744 additions and 374 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
sherpa-onnx/csrc/hypothesis.h
View File

@@ -18,7 +18,7 @@ namespace sherpa_onnx {
struct Hypothesis {
// The predicted tokens so far. Newly predicated tokens are appended.
std::vector<int32_t> ys;
std::vector<int64_t> ys;
// timestamps[i] contains the frame number after subsampling
// on which ys[i] is decoded.
@@ -30,7 +30,7 @@ struct Hypothesis {
int32_t num_trailing_blanks = 0;
Hypothesis() = default;
Hypothesis(const std::vector<int32_t> &ys, double log_prob)
Hypothesis(const std::vector<int64_t> &ys, double log_prob)
: ys(ys), log_prob(log_prob) {}
// If two Hypotheses have the same `Key`, then they contain