-
Hi,
Could you please point out where the gumbel softmax is applied in the code? I am unable to find it.
Thanks
-
https://github.com/vithursant/VAE-Gumbel-Softmax/blob/master/vae_gumbel_softmax.py#L55
You use the logits, however we have to use log(logits) to perform gumbel max trick as stated in [https://timviei…
-
Thank you for your fascinating work! I have several question this paper:
1. Regarding the training of the mask generator, in Formula 9, does the probability represent the binary classification of w…
-
Hi, I find the code in rlattention.py : 'from thumt.layers.gumbel import gumbel_softmax' But in the layers fload, there isn't the Class gumbel ?
-
Could you please share details on gumbel softmax?
How did you incorporate in your project?
Parameters like tau, hard etc?
-
Hello, I have questions on exploration and Gumbel-Softmax.
In the pseudocode, it mentioned initialize random process N for action exploration, which is same in the paper of DDPG. But I have difficu…
-
The equation (1) in this [paper](https://arxiv.org/pdf/1611.01144.pdf) gives the equation of a sampler that samples from a categorical distributions. The advantage of this discrete sampler over the ot…
-
Hi, nice work! I am a bit confused about gumbel softmax. You mention in your paper that, during traininig, gumbel softmax is used. I wonder if it can be replaced by pure softmax (i.e. torch.softmax)? …
-
It doesn’t look like the temperature is annealed in your gumbel softmax. Is there a reason for this as it is not standard? @tkipf
-
I would like to test your code with the transformer architecture in fairseq.
Have you ever tried?
Could you please suggest me the better way to do that?