You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Current issue:
In order to reuse GraphStorm's model components, e.g., InpuLayers and Encoders, users need to create a g(DistGraph) first. This create some barrier for users to build customized models based some built-in compoents.
Suggested solution:
Current GraphStorm components, in effect, only use g(DistGraph) to extract meta information of the graph, such as canonical_etypes or node feature dimensions. Therefore, it would be convenience to just pass those meta information as arguments, which will decouple the binding between model components and g(DistGraph)s.
The text was updated successfully, but these errors were encountered:
Current issue:
In order to reuse GraphStorm's model components, e.g.,
InpuLayer
s andEncoder
s, users need to create a g(DistGraph) first. This create some barrier for users to build customized models based some built-in compoents.Suggested solution:
Current GraphStorm components, in effect, only use g(DistGraph) to extract meta information of the graph, such as canonical_etypes or node feature dimensions. Therefore, it would be convenience to just pass those meta information as arguments, which will decouple the binding between model components and g(DistGraph)s.
The text was updated successfully, but these errors were encountered: