-
# References
+ [Introduction To Autoencoders In Machine Learning](https://youtu.be/NZ97-lFEUq8)
+ [Convolutional autoencoder for image denoising](https://keras.io/examples/vision/autoencoder/)
+ [B…
-
Why do we need pad audio fragment while receiving its mel spec?
`y = torch.nn.functional.pad(y.unsqueeze(1), (int((n_fft-hop_size)/2), int((n_fft-hop_size)/2)), mode='reflect')`
-
Hello there, I want to clarify whether the need for square matrices is strictly enforced. From the paper, I note that
"We turn to an expressive class of sub-quadratic structured matrices called Mo…
-
### 🐛 Describe the bug
I would like to raise a concern about the spectral_norm parameterization.
I strongly believe that Spectral-Normalization Parameterization introduced several versions ago do…
-
## ❓ Questions and Help
Hello,
I try to modify the way convolutions are performed for a conv net (let take ResNet18 as an example) in PyTorch:
Take the first layer. I want to take the FT(Fouri…
-
hello.
thanks for such a cool library.
if you have time, or if possible, please let me know how can I add these operators?
convTranspose, gelu, layerNorm, groupNorm, instanceNorm, swish, mish, loca…
-
HI @tbepler, I am trying to reproduce the results of your paper on the galaxy dataset but unable to exactly achieve those. Could you please share the exact training parameters. I am currently using th…
-
img_numpy = gen_imgs[t,:,:,:].cpu().detach().numpy()
...
psd1D_rec = torch.from_numpy(psd1D_rec).float()
psd1D_rec = Variable(psd1D_rec, requires_grad=True).to(device)
loss_freq = criterion_fr…
-
I believe `conv` is associative:
```julia
julia> x = randn(3); y= randn(4); z=randn(5); conv(conv(x,y),z) ≈ conv(x,conv(y,z))
true
```
so `conv(x,y,z)` is well-defined. Any reason not to support…
-
It seems it should be possible to gpu-accelerate the convolution in DSP.jl, as there are FFTs that are defined for versions of GPUArrays, and elementwise multiplication certainly is. I don't know how …