didadida

Will LM supported?

2

Hi, Is the LM decode(e.g. XX_ngram_LG) will be supported in sherpa & sherpa-ncnn later? because i think adding ngram decode can make the asr app more extensive. Thanks

hi, after testing the project, i find some bad cases: >>> japanese_numbers.to_arabic('作るための四面体をどんどん分割していきましは') [] >>> japanese_numbers.to_arabic('零點零四固定だった') []

[BUG] Lora trianing Missing key in state_dict

2

Feel free to ask any kind of questions in the issues page, but please use English since other users may find your questions valuable. **Describe the bug** Hi, i follow...

Feel free to ask any kind of questions in the issues page, but please use English since other users may find your questions valuable. **Is your feature request related to...

[BUG] the vq result is not well

3

Feel free to ask any kind of questions in the issues page, but please use English since other users may find your questions valuable. **Describe the bug** A clear and...

Feel free to ask any kind of questions in the issues page, but please use English since other users may find your questions valuable. **Describe the bug** A clear and...

Hi, May I ask if i want to train other VQs, such as ar4 or ar8, do i only need to modify the config in vqgan?. Or do i need...

Fail to reproduce the paper result when training from scratch

16

Hi, i try to train the model from scratch, but it fail to reproduce the result in VoiceBank_DEMAND, but the training is unstable and result is bad, can you give...

Feel free to ask any kind of questions in the issues page, but please use English since other users may find your questions valuable. **Describe the bug** Hi, I've been...

Zipvoice: fail to convert onnx to triton for Zipformer model

2

Hi i am trying to convert zipformer to triton, but it failed with error(see below), than i trace the model and found that if we using the list in TTSZipformer,...