Closed Muzzypepper closed 2 years ago
JINHXu
commented
3 years ago Same question here, the Chinese model does not do well with transcribing numbers, if anyone has done related work, experience sharing would be appreciated!
求问有没有大佬搞过中文模型stt识别数字的fine-tune啊，求指教！
Cheers, JX
nshmyrev
commented
3 years ago Hi
We going to retrain Chinese model soon for significantly better accuracy. If you are interesting in cooperation, submit a test set for your tasks so we can debug the accuracy.
JINHXu
commented
3 years ago Hi
We going to retrain Chinese model soon for significantly better accuracy. If you are interesting in cooperation, submit a test set for your tasks so we can debug the accuracy.
Hi!
This is exciting news! Thank you! Should I send it to your gmail?
nshmyrev
commented
3 years ago Yes, you can contact me at nshmyrev@gmail.com or nshmyrev@alphacephei.com.
JINHXu
commented
3 years ago Yes, you can contact me at nshmyrev@gmail.com or nshmyrev@alphacephei.com.
btw can I ask when will the new retrained model be released? We sort of have a deadline here :P Again, thank you very much for the hard work!
nshmyrev
commented
3 years ago btw can I ask when will the new retrained model be released? We sort of have a deadline here :P
No you can't ask for deadlines for free software.
JINHXu
commented
3 years ago btw can I ask when will the new retrained model be released? We sort of have a deadline here :P
No you can't ask for deadlines for free software.
Chill, not asking for a deadline, just wondering if there might be a release date that we can be looking at, so we can consider all our options. Cheers!
nshmyrev
commented
2 years ago Still pending, tracked at #731
I am a newcomer in the field of speech recognition, I want to finetune the chinese model: vosk-model-cn-0.1， but there are few related materials, I hope my seniors can give some guidance。
请问有大佬微调过vosk中文模型 vosk-model-cn-0.1吗，求指教！