Disable test on ARM for CI stability. #6044
Conversation
|
|
||
| using System.Runtime.InteropServices; | ||
|
|
||
| namespace Microsoft.ML.TestFramework.Attributes |
There was a problem hiding this comment.
In Runtime repo we usually use ConditionalFact and we pass there the condition instead of creating a new Fact attribute for different cases. Also, we have there the PlatformDetection which usually contain the properties for different configurations. Did we think to do something similar in machinelearning repo?
There was a problem hiding this comment.
That makes a lot of sense. We don't really have anything like that in machinelearning. The idealistic goal is that there won't be any differences between the different architectures or OS's. When that does happen, it should be the rare exception to the rule. So I think thats why we don't have anything like that. I think using a ConditionalFact makes a lot of sense though. Can you point me to where you do something like that in Runtime and I'll make that change.
There was a problem hiding this comment.
Codecov Report
@@ Coverage Diff @@
## main #6044 +/- ##
==========================================
- Coverage 68.54% 68.53% -0.01%
==========================================
Files 1146 1146
Lines 245450 245451 +1
Branches 25637 25638 +1
==========================================
- Hits 168232 168216 -16
- Misses 70507 70523 +16
- Partials 6711 6712 +1
Flags with carried forward coverage won't be shown. Click here to find out more.
|
ghost
locked as resolved and limited conversation to collaborators
2 participants
The test TestBackAndForthConversionWithAlphaNoInterleave has stability issues on ARM/ARM64. Disabling the test only on those legs for CI stability. Tracked in #6043.