Specify which scopes are removed from TensorBoard main graph by default
See original GitHub issueMigrated from: https://github.com/tensorflow/tensorflow/issues/10421
The heuristics used by TensorBoard to determine which scopes are added to the “main graph” and which are not by default often results in scopes being included that obscure the underlying network structure.
It would be nice to have an additional named parameter for
tf.name_scope
that indicated whether that scope should be removed (or included) in the main graph by default.
Issue Analytics
- State:
- Created 6 years ago
- Reactions:7
- Comments:8
Top Results From Across the Web
TensorFlow - TensorBoard: Graph Visualization
To simplify, variable names can be scoped and the visualization uses this information to define a hierarchy on the nodes in the graph....
Read more >How do I set a name_scope in a graph to be removed from ...
I'm trying to organize my TensorBoard graph such that a certain component is automatically placed on the side when I first initiate ...
Read more >tf.Graph | TensorFlow v2.11.0
A context manager that specifies the default device to use for newly created ops. Raises. RuntimeError, If device scopes are not properly nested....
Read more >TensorBoard Tutorial: Run Examples & Use Logdir - DataCamp
There are several different types of summaries. Here, as you are visualizing only scalars, you can define tf.summary.scalar objects. Furthermore, you can use...
Read more >Save, Load and Inference From TensorFlow Frozen Graph
The major component of pb file is graph structure and also the ... will set its corresponding graph as the default graph globally....
Read more >Top Related Medium Post
No results found
Top Related StackOverflow Question
No results found
Troubleshoot Live Code
Lightrun enables developers to add logs, metrics and snapshots to live code - no restarts or redeploys required.
Start FreeTop Related Reddit Thread
No results found
Top Related Hackernoon Post
No results found
Top Related Tweet
No results found
Top Related Dev.to Post
No results found
Top Related Hashnode Post
No results found
Top GitHub Comments
No progress to report I’m afraid, though I agree this would be useful.
Probably just some kind of extension to
tf.name_scope
. E.g.tf.name_scope(main_graph=True)
that will be inherited to sub scopes and then 1 extra field for the respective protobuf file + some handling logic within tensorboard