-
Notifications
You must be signed in to change notification settings - Fork 615
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
fix the math module for autoray 0.7.0 #6429
Conversation
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.
Looks good.
We will no longer work with autoray < 0.7.0.
Why will autoray<0.7 stop working? We've only registered 2 new functions.
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## master #6429 +/- ##
=======================================
Coverage 99.38% 99.38%
=======================================
Files 447 447
Lines 42429 42441 +12
=======================================
+ Hits 42169 42181 +12
Misses 260 260 ☔ View full report in Codecov by Sentry. |
[sc-76593] |
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.
Only had an optional suggestion to simplify the function registration for non-experimental tensorflow functions, and the idea to remove now redundant pointers to the experimental module that we added earlier.
Thanks for fixing this!
**Context:** Autoray has released a new version. This has caused failures with several tensorflow functions. **Description of the Change:** Patches tensorflow functions that are now longer working. `null.qubit` no longer returns interface numbers unless we have `gradient_method="backprop"`. This is necessary because null qubit was creating tensorflow arrays with the wrong precision. A device does not need to return interface data in any situation other than backprop. That conversion to the interface data type was unnecessary. **Benefits:** No more autoray version pinning. **Possible Drawbacks:** **Related GitHub Issues:**
Context:
Autoray has released a new version. This has caused failures with several tensorflow functions.
Description of the Change:
Patches tensorflow functions that are now longer working.
null.qubit
no longer returns interface numbers unless we havegradient_method="backprop"
. This is necessary because null qubit was creating tensorflow arrays with the wrong precision. A device does not need to return interface data in any situation other than backprop. That conversion to the interface data type was unnecessary.Benefits:
No more autoray version pinning.
Possible Drawbacks:
Related GitHub Issues: