Open innat opened 6 months ago
Hi @innat, this would be a good addition as a Callback! Would you be willing to contribute a PR to add this Profiler Callback?
Yes, I'm willing to contribute. Before that, could you please give some thought on this. Thanks.
Before that, could you please give some thought on https://github.com/keras-team/tf-keras/issues/350#issuecomment-1731852799. Thanks.
There are no profiling plans from our side at this time -- if you want to create this feature, you're welcome to do it, it would be a very valuable addition.
Reopening from here.
Describe the feature and the current behavior/state
In PyTorch-Lightning, you can identify bottlenecks in your code using Profiler API. And I've found it super useful and easy to inspect the pros and cons of the code. This API offers 3 type of inferface, namely
simple
,advance
andpytorch
. I mostly usesimple
that summarize the execution time of medium to high level training + validation + inference API. The saved log files look something as follows.I don't know, in
keras
, if we have anything like this already. By having such high-level functionality inkeras
would really great. It will help to inspect the bottleneck of dataloader, model, callbacks etc, by reporting the time and memory consumption in the system.I looked for it before, discussion.
Will this change the current api? How?
This can be considered as a Callback, I think.
Who will benefit from this feature?
The keras community. Engineers and researchers, both. Mostly engineers in production phase for code optimization.
Contributing