-
-
Notifications
You must be signed in to change notification settings - Fork 153
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
Add rewrite to merge consecutive Join
ed Subtensor
s
#760
Conversation
912d867
to
00da564
Compare
Codecov Report
@@ Coverage Diff @@
## main #760 +/- ##
=======================================
Coverage 78.91% 78.92%
=======================================
Files 152 152
Lines 47657 47698 +41
Branches 10852 10861 +9
=======================================
+ Hits 37609 37646 +37
- Misses 7548 7550 +2
- Partials 2500 2502 +2
|
f5441ae
to
cd6fd7f
Compare
Join
ed Subtensor
s
168a2ce
to
f52bce5
Compare
d455cef
to
675ac90
Compare
675ac90
to
2d17abf
Compare
2d17abf
to
49d4ee0
Compare
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.
A conflict needs to be resolved; otherwise, this looks good to merge.
49d4ee0
to
2df3965
Compare
2df3965
to
81dc59f
Compare
Fixed a wrong assumption that we could perform this rewrite as long as the steps were equal. Also got rid of |
81dc59f
to
e2e54cc
Compare
e2e54cc
to
ccd8360
Compare
I was playing a bit with this kind of rewrites recently, which appeared for me when undoing
at.diff
operations which reduce the first axis index.concatenate(([x[:1], x[1:])) -> x
I am not sure this is a useful rewrite to have in the codebase, let me know what you think.
If this is useful, it will probably still need a bit of cleanup, as I've never worked with SubTensors before.