Skip to content

Conversation

sam-mccall
Copy link
Contributor

The data is about the minimal set that exercises all current features (and #484)

For brevity, it omits some "required" stuff we don't currently use, like the byProgramStructure tree, and the raw{Time,Flops} metrics.

@ioeric

Copy link
Member

@chihuahua chihuahua left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I tried out the demo, and it works nicely. Just a few comments.

@@ -33,6 +33,7 @@
from tensorboard.plugins.profile import profile_demo_data
from tensorboard.plugins.profile import profile_plugin
from tensorboard.plugins.profile import trace_events_pb2
from shutil import copyfile
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Could you replace with import shutil and move it near to import os? Native imports come first, and python style is to import the module, then use it.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I just have this 1 comment, and then lets merge. Thanks!

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done. (Sorry, not sure how I missed this)

@@ -0,0 +1,95 @@
{
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is this expected? I see nothing in the parentheses.

knxthhfajij

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Also, "is overhead, and a good target for optimization" is a dangling half-sentence. I think something might be missing there.

@sam-mccall
Copy link
Contributor Author

Good catch - I left too much out when constructing the data by hand, and the "_worstOp" calculation was crashing.
Added the missing empty 'children' arrays for the leaf nodes, it now looks like this:
image

@chihuahua chihuahua merged commit 7b801ff into tensorflow:master Sep 13, 2017
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants