Alex Barron

Hey, very sorry I'm only just getting back to you all on this. Although I agree it's annoying, it does make sense that there's a big drop off in quality...

Which dataset are you training on? I just uploaded some weights trained on Nancy with r=2, scheduled sampling 0.5 which might be a good starting point. With scheduled sampling 0.5,...

So on the training set it still sounds poor and there's no alignment? I ended up getting better results with r=2 rather than r=5 and so maybe try that or...

No worries at all! Sorry it's been a while. Those clips are from inference on unseen examples (mostly taken from Arctic and the paper examples). It sounds much better during...

Which version of TensorFlow are you running?

It seems like you might have the saved spectrogram with r=5. It should work if you rerun 'preprocess.py nancy' with r=2 (which is now the default in audio.py) and then...

I've been using the CPU Cholesky recently and unfortunately it's quite slow for large matrices. A hybrid CPU/GPU Cholesky with MLX ops is about 2-3x faster than the pure CPU...

I think it's feasible to write a GPU only Cholesky that's at least close to as performant as the above so maybe we don't need to change the pattern. Given...