-
I running inference with the trained vq embedding matrix. In the example ipynb only gives training instructions. I wonder if the edits I made to the source code are valid.
In vqvae.py, quantized …
-
https://github.com/deepmind/sonnet/blob/e097a6678569314f6dd0f0083e91e2cd01aefeff/sonnet/src/nets/vqvae.py#L103
```python
distances = (
tf.reduce_sum(flat_inputs**2, 1, k…
-
Hi, I am trying to build a 2 stage VQ-VAE-2 + PixelCNN as shown in the paper:
"Generating Diverse High-Fidelity Images with VQ-VAE-2" (https://arxiv.org/pdf/1906.00446.pdf).
I have 3 implementation…
-
in_channel=3 -> If I pass in_channel=1, I get an error. I am trying to use it with single channel images. Anyone can help me with this use case?
RuntimeError: Expected 4-dimensional input for 4-di…
-
Hi,
I was going to use your code in one of my projects but the quantization is broken, more specifically [this line](https://github.com/rosinality/vq-vae-2-pytorch/blob/e851d8170709cbe0cdc9521a52f5…
-
Hi
I was comparing your code with the original TensorFlow code made by the authors. It seems you
are not considering q_latent_loss in your Quantize class (check line 101 of https://github.com/deepm…
-
Hi @rosinality, In vqvae.py line no 65,
quantize = input + (quantize - input).detach()
Next this quantize variable we are using input to the decoder.
but from paper we have to give only embeddings…
-
@bshall - First of all, thank you for this implementation. In [this issue](https://github.com/bshall/UniversalVocoding/issues/4), you pointed out that you've generated a sample audio from generated Me…
-
Acc. to the paper the Algorithm goes as follow:
~~~~
....
h_top
-
Hi @pclucas14,
This is the Benji from Rwanda (not Rosman). It's a nice coincidence that we're both looking into VQVAEs!
I've been trying to reproduce the results from "Continuous Relaxation Tra…