-
Notifications
You must be signed in to change notification settings - Fork 5.7k
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
Refine CUDA profiler and delete the test file. #6715
Conversation
python/paddle/v2/fluid/profiler.py
Outdated
"enableonstart 0", | ||
"conckerneltrace", | ||
"active_warps", | ||
"active_warps", |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Why "active_warps" appeared two times, and "cacheconfigrequested" appeared three times?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Fix, Thanks!
config_file = 'nvprof_config_file' | ||
with open(config_file, 'wb') as fp: | ||
fp.writelines(["%s\n" % item for item in config]) | ||
core.nvprof_init(output_file, output_mode, config_file) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Should we check whether output_file
exists? If output_file
has already existed, does this profile result to rewrite the original content of the output_file
?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I do not change the code, all you mentioned can be done by the users when using this interface. Thanks!
Too much options will affect the speed of the program. |
@chengduoZH Thanks for your careful review. I resume the config after a verification. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM+
Fix #6714
Fix #6033