ROCm/jax

Pull Requests Commits

Make Clang use manylinux C++ standard library

charleshofer committed 270 days ago

aad3fa79

Fix C++23 build errors (#257) (#318)

mrodden committed 292 days ago

cefb2ef8

Fix permissions on /tmp (#320)

gabeweisz committed 292 days ago

f7dd3f1c

Update offload ROCm architectures (#311)

Ruturaj4 committed 292 days ago

d152fd0c

Set higher tolerance for failing fp8 tests (#306)

Ruturaj4 committed 292 days ago

3338745d

Fixes for 0.5.0 build (#308)

mrodden committed 292 days ago

4a13859a

Add support for ROCm wheel based install (#283)

mrodden committed 292 days ago

f9654549

Fix and enable RNN tests on JAX 0.5.0 branch (#301)

Ruturaj4 committed 292 days ago

ec5a5fcc

Deprecate unused pre-production gfx versions (#272)

JehandadKhan committed 292 days ago

9237e846

[pallas:triton] Fix atomic min/max lowering for unsigned integers and float types (#264)

draganmladjenovic committed 292 days ago

afa3c0a4

add gfx1101 target (#249) (#254)

JehandadKhan committed 292 days ago

f27e0f2a

Use HIPBLAS_V2 (#222)

Ruturaj4 committed 292 days ago

7087ce90

change latest jaxlib version to 0.5.1 (#298)

Ruturaj4 committed 301 days ago

Verified 1cc753f0

Merge pull request #297 from ROCm/prep-for-v0.5.1

Ruturaj4 committed 301 days ago

Verified c07a0e88

Enable upstream CI on release branches (#296)

charleshofer committed 301 days ago

f56eeaf7

Fix auditwheel version issue (#288)

mrodden committed 301 days ago

c38c23c7

Fix numactl-devel name

charleshofer committed 301 days ago

54ce4c89

Fix numa package

charleshofer committed 301 days ago

c9501cb4

Install numa library

charleshofer committed 301 days ago

79e2c71b

Change CODEOWNERS (#237)

charleshofer committed 301 days ago

065bea83

Use bazel for PR tests (#216)

charleshofer committed 301 days ago

e9168054

Prepare for JAX release 0.5.1

yashk2810 committed 323 days ago

07440f4a

Update XLA dependency to use revision

Google-ML-Automation committed 324 days ago

b0cfcb8e

[sharding_in_types] Add more reshape sharding support

yashk2810 committed 324 days ago

7d3c63ed

[mosaic_gpu] Warmup the kernel when doing CUPTI based profiling

superbobry committed 324 days ago

908ff49e

[pallas:mosaic_gpu] Use `{min,max}imumf` instead of `{min,max}numf`

superbobry committed 324 days ago

74b2e020

Update XLA dependency to use revision

Google-ML-Automation committed 325 days ago

e0e5457c

[sharding_in_types] Add sharding rules for the following primitives:

yashk2810 committed 325 days ago

d695aa4c

Update XLA dependency to use revision

Google-ML-Automation committed 326 days ago

a75bea51

[sharding_in_types] Allow `auto_axes` and `explicit_axes` to take numpy arrays, python scalars.

yashk2810 committed 326 days ago

7c4fe2a7

Older