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
31 Open 58 Closed
31 Open 58 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

jax.jit slows down the code a lot on function with simple array operations and jnp.roll() bug Something isn't working XLA
#24373 opened Oct 17, 2024 by pmocz
4
[ROCM] x64 mode crashes with "redzone_checker with block dimensions: 1024x1x1: hipError_t" AMD GPU Issues pertaining to AMD GPUs (ROCM) bug Something isn't working XLA
#23506 opened Sep 9, 2024 by PhilipVinc
@mrodden
6
JAX segment_sum is two times slower for FP16 inputs than FP32 inputs bug Something isn't working NVIDIA GPU Issues specific to NVIDIA GPUs XLA
#23136 opened Aug 20, 2024 by CloudyDory
10
JIT Compilation Error with Single Vmap on GPU bug Something isn't working GPU Issues that apply to both CUDA and ROCM GPUs XLA
#22865 opened Aug 5, 2024 by bheijden
@jakevdp
1
Vmap of dynamic_update_slice very slow on TPU bug Something isn't working performance make things lean and fast TPU Issues related to compilation for or performance on Google TPUs XLA
#21367 opened May 22, 2024 by j-towns
5
Add "drop" mode to dynamic_update_slice_in_dim enhancement New feature or request P2 (eventual) This ought to be addressed, but has no schedule at the moment. (Assignee optional) XLA
#19627 opened Feb 1, 2024 by jewillco
1
Order of argments to dot_general significantly affects performance bug Something isn't working XLA
#19327 opened Jan 11, 2024 by patrick-kidger
@jakevdp
4
better error message for full remat enhancement New feature or request XLA
#18591 opened Nov 18, 2023 by GallagherCommaJack
2
Associative scan compile very slow for complex numbers bug Something isn't working CPU Issues related to the CPU compiler/runtime P3 (no schedule) We have no plan to work on this and, if it is unassigned, we would be happy to review a PR XLA
#18221 opened Oct 21, 2023 by smorad
3
vmap with scatter_add extremely slow when using xla_gpu_deterministic_ops NVIDIA GPU Issues specific to NVIDIA GPUs performance make things lean and fast XLA
#17844 opened Sep 29, 2023 by BrunoKM
6
5
XLA Check failed: common_utilization <= producer_output_utilization bug Something isn't working GPU Issues that apply to both CUDA and ROCM GPUs needs info More information is required to diagnose & prioritize the issue. XLA
#17730 opened Sep 22, 2023 by pl-fuchs
9
larger integer matrix multiplies are > 20x slower versus floats. bug Something isn't working NVIDIA GPU Issues specific to NVIDIA GPUs XLA
#17688 opened Sep 20, 2023 by ywrt
4
"None of the algorithms provided by cuDNN heuristics worked" for Ampere NV GPUs bug Something isn't working needs info More information is required to diagnose & prioritize the issue. NVIDIA GPU Issues specific to NVIDIA GPUs XLA
#17523 opened Sep 8, 2023 by bentherien
7
lax.conv_transpose takes FOREVER to compile bug Something isn't working NVIDIA GPU Issues specific to NVIDIA GPUs XLA
#17464 opened Sep 6, 2023 by sokrypton
3
slow compile in jax 0.4.6 vs. 0.3.25 bug Something isn't working NVIDIA GPU Issues specific to NVIDIA GPUs XLA
#15290 opened Mar 29, 2023 by sokrypton
7
Increased memory requirement without nn.scan GPU Issues that apply to both CUDA and ROCM GPUs NVIDIA GPU Issues specific to NVIDIA GPUs performance make things lean and fast XLA
#14056 opened Jan 18, 2023 by abhinavgoel95
@hawkinsp
5
Significant Overhead with nn.scan bug Something isn't working NVIDIA GPU Issues specific to NVIDIA GPUs XLA
#13356 opened Nov 22, 2022 by abhinavgoel95
8
pjit + jax.numpy.fft.fftn is not sharding properly enhancement New feature or request NVIDIA GPU Issues specific to NVIDIA GPUs P1 (soon) Assignee is working on this now, among other tasks. (Assignee required) TPU Issues related to compilation for or performance on Google TPUs XLA
#13081 opened Nov 2, 2022 by nvlcambier
2
lax.cond crashes within lax.reduce bug Something isn't working P3 (no schedule) We have no plan to work on this and, if it is unassigned, we would be happy to review a PR XLA
#12615 opened Oct 2, 2022 by soraros
9
Multiplying Nan by False give 0. instead of NaN bug Something isn't working XLA
#12233 opened Sep 6, 2022 by sracaniere
5
jnp.argsort much slower than the numpy version enhancement New feature or request XLA
#10434 opened Apr 23, 2022 by fbartolic
8
vmap + array-indexing is very slow on TPU enhancement New feature or request performance make things lean and fast TPU Issues related to compilation for or performance on Google TPUs XLA
#10378 opened Apr 20, 2022 by ayaka14732
2
XLA: log-indexing should be optimised to indexing-log enhancement New feature or request XLA
#10319 opened Apr 16, 2022 by ayaka14732
3
Attempt to perform FFT requiring batched cufft plan of size > 4 GB causes coredump bug Something isn't working NVIDIA GPU Issues specific to NVIDIA GPUs P1 (soon) Assignee is working on this now, among other tasks. (Assignee required) XLA
#9591 opened Feb 16, 2022 by erdmann
@hawkinsp
5
Efficient argmin/argmax for bool type arrays enhancement New feature or request jax.numpy about JAX's NumPy implementation P3 (no schedule) We have no plan to work on this and, if it is unassigned, we would be happy to review a PR performance make things lean and fast XLA
#9282 opened Jan 21, 2022 by Joshuaalbert
2
ProTip! Type g i on any issue or pull request to go back to the issue listing page.