Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Profiling Tensor Registration and Message Passing #1354

Closed
iamtrask opened this issue Jun 21, 2018 · 2 comments
Closed

Profiling Tensor Registration and Message Passing #1354

iamtrask opened this issue Jun 21, 2018 · 2 comments
Labels
Type: Improvement 📈 Minor improvements not introducing a new feature or requiring a major refactor

Comments

@iamtrask
Copy link
Member

iamtrask commented Jun 21, 2018

see @jvmancuso 's comment elow

@iamtrask iamtrask added Type: Improvement 📈 Minor improvements not introducing a new feature or requiring a major refactor high priority labels Jun 21, 2018
@jvmncs
Copy link
Contributor

jvmncs commented Jun 25, 2018

I think this is part of a larger issue where we should profile operations and tensor sending/receiving to definitively understand where the bottlenecks are. Then we can prioritize optimization efforts analytically/rigorously.

@lucaslopes
Copy link
Contributor

Knows which function is calling register_object might be helpful?

>>> import inspect
>>> def f1(): f2()
>>> def f2():
        curframe = inspect.currentframe()
        calframe = inspect.getouterframes(curframe, 2)
        print 'caller name:', calframe[1][3]
>>> f1()
caller name: f1

Source

@iamtrask iamtrask changed the title Redundant Calls to register_object Profiling Tensor Registration and Message Passing Jul 4, 2018
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Type: Improvement 📈 Minor improvements not introducing a new feature or requiring a major refactor
Projects
None yet
Development

No branches or pull requests

3 participants