Include MoG and ConditionalNorms. #1
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
martinkim0
reviewed
Sep 25, 2023
Comment on lines
+81
to
+82
| cat_dim: int | None = None, | ||
| conditional: bool = False, |
There was a problem hiding this comment.
It currently fails. Can you add something like batch_key (it's the site in mrVI) and ConditionalNorm is only computed on this. It would be 'assay_suspension' for census.
Comment on lines
+60
to
+73
| if self.prior=='mog': | ||
| self.register_buffer( | ||
| "u_prior_logits", torch.ones([mixture_components])) | ||
| self.register_buffer( | ||
| "u_prior_means", torch.randn([n_latent, mixture_components])) | ||
| self.register_buffer( | ||
| "u_prior_scales", torch.zeros([n_latent, mixture_components])) | ||
| elif self.prior=='mog_celltype': | ||
| self.register_buffer( | ||
| "u_prior_logits", torch.ones([n_labels])) | ||
| self.register_buffer( | ||
| "u_prior_means", torch.randn([n_latent, n_labels])) | ||
| self.register_buffer( | ||
| "u_prior_scales", torch.zeros([n_latent, n_labels])) |
There was a problem hiding this comment.
Are these parameters updated during model training?
There was a problem hiding this comment.
Yes they are learnable and are updated in my hands.
| # (n_obs, n_latent) -> (n_obs,) | ||
| kl_div = dist.kl_divergence(posterior, prior).sum(dim=-1) | ||
| if self.prior=='mog' or self.prior=='mog_celltype': | ||
| u = posterior.rsample(sample_shape=(10,)) |
There was a problem hiding this comment.
Should we add the sample shape as a parameter?
There was a problem hiding this comment.
Potentially, in mrVI we fixed it. We should then expose it through model.train.
canergen
commented
Sep 25, 2023
| def forward(self, x, y): | ||
| out = self.bn(x) | ||
| gamma, beta = self.embed(y.long().ravel()).chunk(2, 1) | ||
| out = gamma.view(-1, self.num_features) * out + beta.view(-1, self.num_features) |
There was a problem hiding this comment.
Not sure the .view() thing is needed anymore?
Suggested change
| out = gamma.view(-1, self.num_features) * out + beta.view(-1, self.num_features) | |
| out = gamma * out + beta |
canergen
commented
Sep 25, 2023
| def forward(self, x, y): | ||
| out = self.ln(x) | ||
| gamma, beta = self.embed(y.long().ravel()).chunk(2, 1) | ||
| out = gamma.view(-1, self.num_features) * out + beta.view(-1, self.num_features) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.