docs: fix docstring of selection related methods (#172)

In particular `encode_and_fuse` and `forward` of Selection modules.
ruflab · Oct 11, 2024 · 9b50160 · 9b50160
1 parent b6eba5d
commit 9b50160
Show file tree

Hide file tree

Showing 3 changed files with 12 additions and 6 deletions.
diff --git a/shimmer/modules/global_workspace.py b/shimmer/modules/global_workspace.py
@@ -296,7 +296,8 @@ def encode_and_fuse(
 
         Args:
             x (`LatentsDomainGroupsT`): the input domain representations.
-            selection_scores (`Mapping[str, torch.Tensor]`):
+            selection_module (`SelectionBase`): selection module to use to obtain
+                selection scores.
 
         Returns:
             `dict[frozenset[str], torch.Tensor]`: the GW representations.

diff --git a/shimmer/modules/gw_module.py b/shimmer/modules/gw_module.py
@@ -298,8 +298,8 @@ def encode_and_fuse(
 
         Args:
             x (`LatentsDomainGroupT`): the input domain representations
-            selection_score (`Mapping[str, torch.Tensor]`): attention scores to
-                use to encode the reprensetation.
+            selection_module (`SelectionBase`): selection module to use to obtain
+                selection scores.
 
         Returns:
             `torch.Tensor`: The merged representation.

diff --git a/shimmer/modules/selection.py b/shimmer/modules/selection.py
@@ -38,6 +38,8 @@ def forward(
 
         Args:
             domains (`LatentsDomainGroupT`): Group of unimodal latent representations.
+            encodings_pre_fusion (`LatentsDomainGroupT`): pre-fusion domain latent
+                representation.
 
         Returns:
             `dict[str, torch.Tensor]`: for each domain in the group, the fusion
@@ -75,7 +77,8 @@ def forward(
 
         Args:
             domains (`LatentsDomainGroupT`): input unimodal latent representations
-            gw_state (`torch.Tensor`): the previous GW state
+            encodings_pre_fusion (`LatentsDomainGroupT`): pre-fusion domain latent
+                representation.
 
         Returns:
             `dict[str, torch.Tensor]`: whether the domain is selected for each input
@@ -105,7 +108,8 @@ def forward(
 
         Args:
             domains (`LatentsDomainGroupT`): input unimodal latent representations
-            gw_state (`torch.Tensor`): the previous GW state
+            encodings_pre_fusion (`LatentsDomainGroupT`): pre-fusion domain latent
+                representation.
 
         Returns:
             `dict[str, torch.Tensor]`: whether the domain is selected for each input
@@ -281,7 +285,8 @@ def forward(
 
         Args:
             domains (`LatentsDomainGroupT`): Group of unimodal latent representations.
-            encodings (`LatentsDomainGroupT`): Group of pre-fusion encodings.
+            encodings_pre_fusion (`LatentsDomainGroupT`): pre-fusion domain latent
+                representation.
 
         Returns:
             `dict[str, torch.Tensor]`: the attention scores for each domain in the