jakevdp
Follow
Create your own GitHub profile
Sign up for your own profile on GitHub, the best place to host code, manage projects, and build software alongside 50 million developers.
Sign upPinned
1,600 contributions in the last year
Contribution activity
November 2020
Created 30 commits in 1 repository
Created a pull request in google/jax that received 5 comments
PRNGSeed: ensure JIT invariance for valid inputs.
This is a less drastic version of #4821 that aims only for JIT invariance for jax.random.PRNGKey, leaving dtype invariance breakages untouched. The…
+32
−16
•
5
comments
Opened 31 other pull requests in 1 repository
google/jax
26
merged
3
open
2
closed
- Fix random doc
- Support 'where' argument in jax.numpy reductions
- Make jnp.gradient() call signature match np.gradient
- Cleanup: remove obsolete TODOs
- Support 'initial' argument in jax.numpy reductions
- Increase test coverage for reductions.
- jax.numpy: match numpy arguments for reductions
- Add initial changelog for jax v0.2.7
- update version & CHANGELOG for jax v0.2.6
- Remove noisy warnings for var() of int
- Make dtype an explicit argument to nan-reductions
- fix mypy error
- Fix mypy
- Support shape parameter in jax.numpy.*_like()
- lax_numpy: fix handling of out keyword in wrapped methods
- Update in1d() TODO
- Fix dtypes.is_weakly_typed
- TST: add explicit tests of PRNGKey seed handling.
- Deprecation: Raise a TypeError when indexing with a non-tuple sequence
- DOC: update webpage copyright year to 2020
- cleanup: remove unused test arg
- WIP: propagate weak types through unary and binary ops
- PRNGKey: more careful handling of typecasting for various inputs.
- DOC: fix generating code in type_promotion.rst
- DOC: add lattice to type promotion doc
- Some pull requests not shown.
Reviewed 5 pull requests in 1 repository
google/jax 5 pull requests
Created an issue in google/jax that received 2 comments
jit fails for large uint64 inputs
Repro: import jax.numpy as jnp from jax import jit, random, config config.update('jax_enable_x64', True) x = jnp.iinfo(jnp.uint64).max print(jnp.ui…
2
comments