Matthew Honnibal
Matthew Honnibal
The issue is the Python requirement. We've had to stop building for Python 3.8, as our dependencies no longer support it.
I've been trying to get wheels built against numpy v2 for `cython-blis` but I'm having trouble with our wheel-building machinery. I'm looking at just setting the upper pin for now...
Just published a new version with the updated pin. This should fix the installation. I'll still work on getting Thinc built against numpy v2 though.
I believe this should be resolved now. Please reopen (or open a new issue) if there's continued problems. Note that it's best to install into a clean environment for this...
Current versions of spaCy now support a wider numpy compatibility range, which should resolve this.
> But, current CuPy Cython-level interface is unstable and non-documented. > Cython has no access control. If we open it, user may use all method that includes internal method. >...
There's no way to support both numpy v1 and numpy v2 in the same release. You'll need to pin your spaCy dependency to have an upper-bound that uses numpy v1.
This should be resolved now. numpy v2 is specified as a build-time dependency, while runtime is compatible with both.
Is there a reason you're trying to install this particular version? It's quite old