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
We see that the Percentage column doesn't sum to 100, and indeed, model_forward in contained inside each run_training_batch thus its time is partial to run_training_batch time.
But what about model_backward? is it contained inside model_forward? I assume no, but it there is the ability to see the profiler in a "graph" like structure, where you can understand what operations are "leaf" nodes, or part-of some other operation, it will really ease up the analysis.
Example (taken from pytorch docs):
The text was updated successfully, but these errors were encountered:
Looking at the given profiler status:
We see that the
Percentage
column doesn't sum to 100, and indeed,model_forward
in contained inside eachrun_training_batch
thus its time is partial torun_training_batch
time.But what about
model_backward
? is it contained insidemodel_forward
? I assume no, but it there is the ability to see the profiler in a "graph" like structure, where you can understand what operations are "leaf" nodes, or part-of some other operation, it will really ease up the analysis.Example (taken from pytorch docs):

The text was updated successfully, but these errors were encountered: