Partially specified NodeDef.attr #84
Labels
enhancement
New feature or request
meta graph
This issue involves the meta graph objects
TensorFlow
This issue involves the TensorFlow backend
It might be worth investigating a way to partially specify the
NodeDef.attr
dict
, especially since complete removal/reassignment causes problems for constants (the Numpy constant value is held in that map). In other words, it would be like saying "NodeDef.attr
is adict
that at least contains a"data"
key."A nice way of doing this could make use of a partially specified
cons
map: i.e. an object likecons(('data', 1), ..., var())
that representsOrderedDict(('data': 1), ...)
. We would need a helper class to provide the standarddict
interface and a means of stating that thecdr
must ultimately be anOrderedDict
; otherwise,cons
will already work in this capacity.Originally posted by @brandonwillard in #35 (comment)
The text was updated successfully, but these errors were encountered: