jax
Add ROCm benchmark workflow for MaxText
#755
Closed

Add ROCm benchmark workflow for MaxText #755

psanal35 wants to merge 50 commits into amd-main from add-rocm-model-benchmarks
psanal35
psanal35 psanal35 force pushed from 6e71c3d2 to 89e0f5a6 30 days ago
psanal35 psanal35 force pushed from 89e0f5a6 to a7456912 30 days ago
psanal35 psanal35 force pushed from a7456912 to 7270aea6 30 days ago
psanal35 psanal35 force pushed from 7270aea6 to 96a49665 30 days ago
psanal35 psanal35 force pushed from 96a49665 to 5e8fd901 30 days ago
mminutoli
mminutoli commented on 2026-04-23
psanal35 psanal35 force pushed from 5e8fd901 to eeaa06c7 30 days ago
psanal35 psanal35 force pushed from eeaa06c7 to b98b1114 30 days ago
psanal35 psanal35 force pushed from b98b1114 to 86c526ce 29 days ago
psanal35 psanal35 force pushed from 86c526ce to 2c60dd74 29 days ago
psanal35 psanal35 force pushed from 2c60dd74 to fff9d8bd 29 days ago
psanal35 psanal35 force pushed from fff9d8bd to 517bb143 29 days ago
psanal35 psanal35 force pushed from 517bb143 to eedae290 19 days ago
psanal35 psanal35 force pushed from eedae290 to 6f08c2a6 19 days ago
charleshofer
charleshofer requested changes on 2026-05-05
charleshofer Remove nvidia_wheel_versions
43c0570e
charleshofer Make jaxlib targets visible
bcef89c3
charleshofer hipblas typedef fix
733b7bf8
charleshofer No GPU fail
793d3127
mminutoli Wrap HIP inline functions in anonymous namespaces in vendor.h
e3ad0ecb
dsicarov-amd SWDEV-512768 - Replace hipGetLastError with hipExtGetLastError
a831ef20
charleshofer Add shared utility function get_rocm_version to test_util.py
58249a4a
phambinhfin Fix hipSparse CSR algorithm mappings for ROCm 7
e587f903
phambinhfin Fix v_pages quantization and adjust test params for ROCm compatibilit…
80899473
Arech8 Address LLVM assertion failure due to a multithreaded use. Update .gi…
d9e7020e
Arech8 Add skip of test_is_finite() on Cuda (#565)
42a3be64
AratiGanesh Add rocm test requirements file (#570)
544c6d4a
charleshofer Let the unit tests use build.py for setting up Bazel commands for uni…
4673584d
gulsumgudukbay adding abort logic to rocm/jax (#590)
1c79814a
phambinhfin Skip is_finite tests on ROCm (not in Triton lowering for jax 0.8.0) (…
9b5d7088
phambinhfin Fix shared memory limit check for ROCm in test_dot (#596)
82bf13e1
magaonka-amd Fix Numpy signatures test (#598)
ad47e174
Ruturaj4 fix merge arts
3b3b31cf
gulsumgudukbay Enable RngShardingTests (#644)
8a9adefd
mminutoli Enable test_variadic_reduce_window on ROCm (#647)
4eb74735
magaonka-amd Skip sparse tests on ROCm due to hipSPARSE issue (#652)
f360e135
magaonka-amd Update sparse test skip messages in v0.8.2 (#653)
81842f41
magaonka-amd Skip sparse tests on ROCm due to hipSPARSE issue (#652)
489fcf60
magaonka-amd Update sparse test skip messages in v0.8.2 (#653)
c2ea7b45
magaonka-amd Skip sparse tests on ROCm due to hipSPARSE issue (#652)
82a1e816
magaonka-amd Update sparse test skip messages in v0.8.2 (#653)
2c12a03c
AratiGanesh Enable testMultivariateNormalSingularCovariance on ROCm (#666)
5c681b10
gulsumgudukbay Update Skip Reason Outputs (#663)
3757b64d
magaonka-amd Skip sparse tests on ROCm due to hipSPARSE issue (#652)
7ec9fe0f
magaonka-amd Update sparse test skip messages in v0.8.2 (#653)
411d4faa
magaonka-amd Skip testCudaArrayInterfaceOnNonCudaFails on ROCm platform (#677)
130ca422
magaonka-amd Skip sparse tests on ROCm due to hipSPARSE issue (#652)
837654d2
magaonka-amd Update sparse test skip messages in v0.8.2 (#653)
2b8c7fe3
magaonka-amd Skip sparse tests on ROCm due to hipSPARSE issue (#652)
d0a11b30
magaonka-amd Update sparse test skip messages in v0.8.2 (#653)
3e79165b
magaonka-amd Remove 'mean' from unsupported params for jnp.var (#689)
9d4fce1a
AratiGanesh Skipping testEighTinyNorm due to hipSolver issues (#697)
7601b868
gulsumgudukbay Abort detection CI workflow (#688)
c21c3b8d
gulsumgudukbay Abort-Detection: Fix halt-for-connection input (#712)
3f5828e9
WBobby fix: add rocm_sysdeps/lib to wheel RUNPATH (#737)
6283bbcf
psanal35 Temporarily disable the cron trigger for the cont. wheel tests workflow
2422c280
mminutoli mminutoli force-pushed the amd-main branch from ddcf4f41 to 2422c280 17 days ago
psanal35 Add placeholder for nightly benchmark workflow (#768)
8d4fbef1
psanal35 Rename the benchmarks workflow consistently (#770)
eb58ba29
psanal35 Add ROCm benchmark workflow for MaxText
2483b469
psanal35 psanal35 force pushed from 6f08c2a6 to dd197185 13 days ago
psanal35 psanal35 force pushed from dd197185 to 97c69f97 13 days ago
psanal35 psanal35 force pushed from 97c69f97 to 2483b469 13 days ago
psanal35 psanal35 force pushed from 703bea1e to 7101e890 13 days ago
psanal35 psanal35 force pushed from 7101e890 to 4a5afeef 13 days ago
psanal35 Resolve latest MaxText Transformer Engine wheel from ROCm MaxText rel…
1d25b87d
psanal35 psanal35 force pushed from 4a5afeef to 1d25b87d 13 days ago
psanal35 psanal35 force pushed from 1f99d5e4 to 81fc20fb 13 days ago
psanal35 psanal35 force pushed from 81fc20fb to d95ef011 13 days ago
psanal35 psanal35 force pushed from d95ef011 to 555e4c29 13 days ago
psanal35 psanal35 force pushed from 555e4c29 to 84053303 13 days ago
psanal35 Load MaxText ROCm benchmark configs and requirements from ROCm/maxtext
24fb105b
psanal35 psanal35 force pushed from 84053303 to 24fb105b 13 days ago
psanal35 psanal35 force pushed from b09f0181 to 8604c5e5 12 days ago
psanal35 psanal35 force pushed from 8604c5e5 to a52d98cf 12 days ago
psanal35 psanal35 force pushed from a52d98cf to d02886b5 12 days ago
psanal35 psanal35 force pushed from d02886b5 to 08927951 12 days ago
psanal35 psanal35 force pushed from 08927951 to cd5ad81f 12 days ago
psanal35 Revisit ROCm benchmark results and run-manifest collection
34b2d0f8
psanal35 psanal35 force pushed from cd5ad81f to 34b2d0f8 12 days ago
psanal35 Revisit ROCm artifact upload to S3 for reusability
37fe3f7a
psanal35 Remove TE installation to keep the model lightweight
9183f204
psanal35 Update benchmark target scripts for more generic use cases
182dd2a8
mminutoli mminutoli force-pushed the amd-main branch from eb58ba29 to fb485937 4 days ago
psanal35
psanal35 psanal35 closed this 4 days ago
psanal35 psanal35 deleted the add-rocm-model-benchmarks branch 4 days ago

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone