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
Hello,
I followed the example from documentation (https://jax.readthedocs.io/en/latest/profiling.html) to set up profiling of my code with Perfetto. I used jax.profiler.annotate_function() (https://jax.readthedocs.io/en/latest/_autosummary/jax.profiler.annotate_function.html#jax.profiler.annotate_function) to annotate one of the functions. I then tested the code for varying problem sizes and I found out that for larger problems where the function takes a very small fraction of the overall runtime, the cusotm trace was not visible anymore in Perfetto. Is this a bug or a known behaviour of the tracer? Also, I am considering switching to tensorboard but I really don't want to have to install tensorflow in my environment just to be able to use it. Are the differences between the two purely aesthetic or does tensorboard provide more functionalities?
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
-
Hello,
I followed the example from documentation (https://jax.readthedocs.io/en/latest/profiling.html) to set up profiling of my code with Perfetto. I used
jax.profiler.annotate_function()
(https://jax.readthedocs.io/en/latest/_autosummary/jax.profiler.annotate_function.html#jax.profiler.annotate_function) to annotate one of the functions. I then tested the code for varying problem sizes and I found out that for larger problems where the function takes a very small fraction of the overall runtime, the cusotm trace was not visible anymore in Perfetto. Is this a bug or a known behaviour of the tracer? Also, I am considering switching to tensorboard but I really don't want to have to install tensorflow in my environment just to be able to use it. Are the differences between the two purely aesthetic or does tensorboard provide more functionalities?Beta Was this translation helpful? Give feedback.
All reactions