vell001
vell001
Thanks a lot~ I have anther question, your model cannot split very well, when the song with some special drumbeats, like this [test001.mp3.zip](https://github.com/madebyollin/acapellabot/files/1051965/test001.mp3.zip) My question is can I solve this...
ok, thanks, please let me know if you have any progress, I'll try to training based on your model with more special style data thanks again~
后面不断在升级了，这个demo应该也不适用了，还是参考官方文档吧 [https://bugly.qq.com/docs/user-guide/instruction-manual-android-upgrade/](https://bugly.qq.com/docs/user-guide/instruction-manual-android-upgrade/)
https://www.github.com/farizrahman4u/recurrentshop
这是bug吧？同遇到了，定位了半天问题，求修复啊