You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Obviously not just yet, but once NumPy 2.0 is released and stable, we should drop support for NumPy 1.26 as a backend. That would allow us to remove all the compatibility code and only have code related to strictness.
The actual backend in array-api-strict doesn't really matter that much, but this would limit the usability for libraries that use NumPy internally in addition to the array API, so we should only do this once the ecosystem seems generally caught up on NumPy 2.0 support.
Also I'm not sure if it's worth trying to implement some of the new dlpack features without direct numpy support for them (I need to do some more investigation), so we may want to just always keep this package pinned to the latest numpy. See #35
NumPy 2.0 is out. This is low priority right now since everything works as is, but we definitely could simplify a lot of the code if we required NumPy 2.0.
Looks like this was premature. I reverted the NumPy 2.1 dependency in array-api-strict 2.1.3. Some packages don't support NumPy 2.0 yet. We should definitely do this because it will let use remove a lot of the compat code here (which really should only be existing in array-api-compat), but we can wait a little while longer.
Activity
asmeurer commentedon May 1, 2024
Also I'm not sure if it's worth trying to implement some of the new dlpack features without direct numpy support for them (I need to do some more investigation), so we may want to just always keep this package pinned to the latest numpy. See #35
asmeurer commentedon Jun 26, 2024
NumPy 2.0 is out. This is low priority right now since everything works as is, but we definitely could simplify a lot of the code if we required NumPy 2.0.
Require NumPy >= 2.1
asmeurer commentedon Nov 8, 2024
Looks like this was premature. I reverted the NumPy 2.1 dependency in array-api-strict 2.1.3. Some packages don't support NumPy 2.0 yet. We should definitely do this because it will let use remove a lot of the compat code here (which really should only be existing in array-api-compat), but we can wait a little while longer.