Skip to content

Issues: jax-ml/jax

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Clear current search query, filters, and sorts
24 Open 174 Closed
24 Open 174 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

⚠️ Nightly upstream-dev CI failed ⚠️ CI
#24544 opened Oct 26, 2024 by github-actions bot
@dfm
2
Update pre-commit workflow to cache on jax version pull ready Ready for copybara import and testing
#24543 opened Oct 26, 2024 by dfm Loading…
@dfm
Update jaxlib GPU FFI handlers to use templated typed_data method rather than explicit casting.
#24534 opened Oct 25, 2024 by copybara-service bot Loading…
@dfm
Allow dot algorithms in default_matmul_precision config
#24510 opened Oct 24, 2024 by dfm Loading…
@dfm
hlo_helpers.dtype_to_ir_type is not complete comparing with jax.interpreters.mlir.dtype_to_ir_type enhancement New feature or request
#24509 opened Oct 24, 2024 by aws-qieqingy
@dfm
7
Run dot algorithm tests with PJRT plugin pull ready Ready for copybara import and testing
#24480 opened Oct 23, 2024 by dfm Loading…
@dfm
Clean up conditional lowering of tan after JAX v0.4.34 release. pull ready Ready for copybara import and testing
#24405 opened Oct 19, 2024 by dfm Loading…
@dfm
Migrate CUDA FFI example to GitHub Actions using hosted GPU runner
#24190 opened Oct 8, 2024 by dfm Loading…
@dfm
4
[WIP] skip zeroing weights when computing weight matrix for scale_and_translate
#24111 opened Oct 4, 2024 by barney-s Loading…
@dfm
scale_and_translate with output_size and scale returning zero values for last row and column bug Something isn't working
#24106 opened Oct 3, 2024 by barney-s
@dfm
1
BUG: return one covariance matrix per rhs in polyfit
#24074 opened Oct 2, 2024 by scottstanie Loading…
@dfm
6
jnp.polyfit with cov=True only returns one covariance matrix for multiple right hand sides bug Something isn't working
#24073 opened Oct 2, 2024 by scottstanie
@dfm
Re-enable cudnn_fusion_test on A100. pull ready Ready for copybara import and testing
#24069 opened Oct 2, 2024 by sergachev Loading…
@dfm
4
Orthogonal Initializer raises gpusolverDnCreate(&handle) failed: cuSolver internal error bug Something isn't working
#23616 opened Sep 13, 2024 by brianorbrain
@dfm
3
CuSolver: Switch to 64 bit api to allow for eigh on matrices > than 26732x26732 enhancement New feature or request
#23413 opened Sep 3, 2024 by PhilipVinc
@dfm
6
Improve CuSolver errors diagnostics enhancement New feature or request
#23410 opened Sep 3, 2024 by PhilipVinc
@dfm
8
Port jaxlib's triton_call target to the FFI.
#23249 opened Aug 26, 2024 by copybara-service bot Loading…
@dfm
Port all custom calls to FFI enhancement New feature or request
#23056 opened Aug 14, 2024 by dfm
14 of 20 tasks
@dfm
Accessing incoming gradient and activations in custom_vjp enhancement New feature or request
#22993 opened Aug 12, 2024 by yidingjiang
@dfm
4
When caching is enabled, also enable XLA caching features as well pull ready Ready for copybara import and testing
#22899 opened Aug 6, 2024 by trevor-m Loading…
@dfm
17
[WIP] Add a custom_ad interface to allow simultaneous customization of JVP and VJP rules
#22457 opened Jul 15, 2024 by dfm • Draft
@dfm
Larger matrix diagonalization returns OverflowError enhancement New feature or request
#20904 opened Apr 23, 2024 by OMalenfantThuot
@hawkinsp @dfm
Defining both custom_vjp and custom_jvp question Questions for the JAX team
#3996 opened Aug 8, 2020 by thisiscam
@dfm
11
Support (nonsymmetric) np.linalg.eig on GPU enhancement New feature or request NVIDIA GPU Issues specific to NVIDIA GPUs P3 (no schedule) We have no plan to work on this and, if it is unassigned, we would be happy to review a PR
#1259 opened Aug 28, 2019 by clemisch
@dfm
31
ProTip! Mix and match filters to narrow down what you’re looking for.