-
Hi,
Thanks for sharing this great work. I have just one question:
During the training of all models, did you face a problem of mismatched energy levels between the target and generated speech wave…
-
### Description
I am having trouble training models that don't rely on an f0 estimate from the Crepe pitch estimator. In my tests, whenever fundamental frequency estimation is part of the different…
-
Explore size reductions on EA files after DL3 is implemented.
1. Version 0 of DL3: 10GByte per combined EA file. That is huge in comparison with the
-
Hi again,
I would like to ask for your experience with this model. In my case, I found the model converges very fast, even with 10K steps training (batch size 64). I tried significantly reducing mo…
-
Dear Pranay,
I am working on generative models for musical sound synthesis. As other related works, the optimization of the generation uses spectrogram reconstructions (e.g. multiscale spectral los…
-
Dear guys,
I recognize something that to me doesn't make any sense. Perhaps the function `spec()` needs to be checked again.
The problem is the following:
I load a dataset using:
`A1
-
Hello,
While running a simulation with oskarpy using a python script, the simulation is taking only one sky model. The inputs are given as :
`sim.set_sky_model('point_source.osm')`
`sim.set_sky_mod…
-
Hi, so far I had good results switching to your library for computing spectral reconstruction losses for raw waveform generation !
However, using CQT2010v2 does 10 prints "downsample_factor = 4" …
-
**Submitting author:** @zingale (M. Zingale)
**Repository:** https://github.com/amrex-astro/Castro
**Version:** 20.07
**Editor:** @eloisabentivegna
**Reviewers:** @kegiljarhus, @bonh
**Managing EiC:*…
-
I noticed that the PyTorch version is not just the same as the paper, and have some strange points.
For example:
1. `self.wganloss = WGANLoss()` is never used, which is used in the paper
2. The `…