Karel Vesely
3f2a17ef47
Fixes issue #535 , fix hexa 1-char tokens in ASR output. ( #550 )
...
- Avoid output like : `[' K', '<0x64>', '<0x79>', 'ť', ' a', '<0x75>',
'to', 'bu', '<0x73>', '<0x75>', ... ]` with regular 500 BPE units.
- Don't rewrite 1-char tokens in range [ 0x20 (space) .. 0x7E (tilde) ]
2024-01-26 19:23:20 +08:00
chiiyeh
3bb3849ec5
add blank_penalty for offline transducer ( #542 )
2024-01-25 15:00:09 +08:00
Fangjun Kuang
e215d0c39a
Fix Byte BPE string results for Python. ( #512 )
...
It ignores invalid UTF8 strings.
2024-01-03 16:03:24 +08:00
Wei Kang
47184f9db7
Refactor hotwords，support loading hotwords from file ( #296 )
2023-09-14 19:33:17 +08:00
Fangjun Kuang
debab7c091
Add two-pass speech recognition Android/iOS demo ( #304 )
2023-09-12 15:40:16 +08:00
Wei Kang
8562711252
Implement context biasing with a Aho Corasick automata ( #145 )
...
* Implement context graph
* Modify the interface to support context biasing
* Support context biasing in modified beam search; add python wrapper
* Support context biasing in python api example
* Minor fixes
* Fix context graph
* Minor fixes
* Fix tests
* Fix style
* Fix style
* Fix comments
* Minor fixes
* Add missing header
* Replace std::shared_ptr with std::unique_ptr for effciency
* Build graph in constructor
* Fix comments
* Minor fixes
* Fix docs
2023-06-16 14:26:36 +08:00
keanu
1a1b9fd236
RNNLM model support lm_num_thread and lm_provider setting ( #173 )
...
* rnnlm model inference supports num_threads setting
* rnnlm params decouple num_thread and provider with Transducer.
* fix python csrc bug which offline-lm-config.cc and online-lm-config.cc arguments problem
* lm_num_threads and lm_provider set default values
---------
Co-authored-by: cuidongcai1035 <cuidongcai1035@wezhuiyi.com >
2023-06-12 15:51:27 +08:00
keanu
9c017c2ccb
rnnlm model inference supports num_threads setting ( #169 )
...
Co-authored-by: cuidongcai1035 <cuidongcai1035@wezhuiyi.com >
2023-06-07 09:32:27 +08:00
Fangjun Kuang
86017f9833
Add RNN LM rescore for offline ASR with modified_beam_search ( #125 )
2023-04-23 17:15:18 +08:00
Fangjun Kuang
423d89e9a5
Support paraformer. ( #95 )
2023-03-28 17:59:54 +08:00
Fangjun Kuang
dffb0fd43c
Refactor offline recognizer. ( #94 )
...
* Refactor offline recognizer.
The purpose is to make it easier to support different types of models.
2023-03-27 14:59:40 +08:00
