-
Hi,
I've implemented the model search on my own since it is not available yet in your github repo.
Here is my SuperNet-Code of the encoder for the XLarge Model. Note, that I ommited the classifi…
maaft updated
4 years ago
-
I'm not seeing a bunch of bisectors; am I missing something?
import tensorflow_probability as tfp
tfd = tfp.distributions
tfb = tfp.bijectors
dir(tfb)
yields:
['AbsoluteValue',
'Af…
-
-
![Group 81](https://user-images.githubusercontent.com/17490886/69956678-fb388980-1543-11ea-86ac-b18b8c31f1e7.png)
-
Thank you for making the search code publicly accessible, it is really a good material for Nas researchers. After reading the paper and reviewing the search code, there a few points which I may not cl…
-
### My situation
I want to apply bits-back coding to a VAE with categorical latent variables. Specifically, each latent variable is distributed according to a categorical distribution with 256 possib…
-
## Context
We would like to add `torch::nn::functional::gumbel_softmax` to the C++ API, so that C++ users can easily find the equivalent of Python API `torch.nn.functional.gumbel_softmax`.
##…
yf225 updated
4 years ago
-
You are using argmax between the first encoder and second decoder. How is the back propagation happening since argmax is non-differentiable ?
-
hi, may I know which paper(s)/method the discrete SAC is based upon?
from my understanding there are 3 main implementations of discrete SAC:
- Gumbel Softmax
- KL Divergence
- Petros Christo…
-
Currently, options are represented with a two-dimensional array but one should suffice. For example, change option representation from [[1,0,0],[1,0,0]] to [1,0,0]. Gumbel softmax and everything else …