Make simple inference from #1 with sample_rate configurable

Open hahunavth opened this issue 2 years ago • 3 comments

May 17 '23 03:05 hahunavth

@awarebayes @hahunavth Thank you for providing an inference script. is it possible to also provide a pretrained model for testing?

Nov 22 '23 06:11 HeChengHui

@HeChengHui Sorry for the late response. The author of this repository has provided the pretrained model here. I tried it out and found that the audio quality after denoising is also quite good.

Nov 23 '23 16:11 hahunavth

@hahunavth sorry if this is the wrong place to ask as i dont see an issues tab. The paper mentions a few time that this is a causal model, but both denoise.py and denoise_simple.py seems to take audio file as input. How do i get this model to work with causal/audio stream?

Nov 24 '23 05:11 HeChengHui