Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

rename terms in low rank mvn autoguide #1600

Merged
merged 2 commits into from
Dec 1, 2018
Merged

rename terms in low rank mvn autoguide #1600

merged 2 commits into from
Dec 1, 2018

Conversation

jpchen
Copy link
Member

@jpchen jpchen commented Dec 1, 2018

@@ -521,15 +521,15 @@ class AutoLowRankMultivariateNormal(AutoContinuous):
guide = AutoLowRankMultivariateNormal(model, rank=10)
svi = SVI(model, guide, ...)

By default the ``D_term`` is initialized to 1/2 and the ``W_term`` is
intialized randomly such that ``W_term.matmul(W_term.t())`` is half the
By default the ``diagonal_term`` is initialized to 1/2 and the ``cov_factor`` is
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

...the cov_diag is initialized to...

pyro.param("auto_W_term", torch.randn(latent_dim, rank)))
pyro.param("auto_D_term", torch.randn(latent_dim).exp()),
pyro.param("auto_cov_factor", torch.randn(latent_dim, rank)))
pyro.param("auto_cov_diagonal", torch.randn(latent_dim).exp()),
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It looks like these are named cov_factor and cov_diag upstream.
https://pytorch.org/docs/master/_modules/torch/distributions/lowrank_multivariate_normal.html
Can we match those as auto_cov_factor and auto_cov_diag?

@fritzo fritzo merged commit 27e9d1a into dev Dec 1, 2018
@martinjankowiak martinjankowiak deleted the rename-lrmvn branch December 2, 2018 16:23
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants