num_groups_in_drop_band 和训练的batch size的关系

Audio-WestlakeU / FullSubNet

PyTorch implementation of "FullSubNet: A Full-Band and Sub-Band Fusion Model for Real-Time Single-Channel Speech Enhancement."

https://fullsubnet.readthedocs.io/en/latest/

MIT License

551 stars 155 forks source link

Open xf739645524 opened 3 years ago

xf739645524 commented 3 years ago

我可以使用batch size为8 (大于1设置), 然后 num_groups_in_drop_band 设置为1, 这样可以同时训练多个batch size音频不, 加快训练速度