Nathan Goldbaum
Nathan Goldbaum
Are there other pain spots where it would be useful for me to look? Happy to do so: unblocking Numba on 3.14t is high on our team's priority list right...
> Lastly are we all in agreement that this PR shows that llvmlite is very, very likely to work just fine ™️ in the free-threading/no-gil context? If someone is using...
Here's a passing CI run from my fork: https://github.com/ngoldbaum/python-lz4/actions/runs/19049344029
@jonathanunderwood I see you did another release yesterday. It'd be really nice to get this PR merged and ship another release with this. Right now there's no cp314t wheel, which...
Ping @jonathanunderwood - I'd really appreciate having a maintainer look this over. Let me know if you'd like anything from me to help get this over the line.
I ended up using this query: https://gist.github.com/ngoldbaum/57933c64dec7575e9aa1134690ebba76 This works for my purposes, although it’s imperfect because it uses sampling to avoid high costs of processing the full dataset.
I tested this with spaCy built from-source using all of the latest versions of our WIP work on the full stack with Python 3.13. See [this gist](https://gist.github.com/ngoldbaum/a07fd32e12c8004148d666336b3f15c3) for my full...
@honnibal does the spaCy test suite passing with this PR and the discussion above assuage your concerns?
The old cython version is because of https://github.com/nascheme/cpython_sanity/issues/9 so this might be fixed already.
Can you try again with numpy 1.26.4? Numpy 1.26.3 had a number of f2py-related fixes.