jax
[Test do not merge] Testing build wheels
#722
Closed

[Test do not merge] Testing build wheels #722

alekstheod wants to merge 7513 commits into main from amd-main
alekstheod
dependabot[bot] Bump actions/checkout from 6.0.0 to 6.0.1
523b667b
yashk2810 Propagate input sharding to output in full_like if it's fully replica…
3b6fd04e
mattjj [hijax] add is_vjp bit to LinearizeTrace
79f3519e
Google-ML-Automation Merge pull request #35285 from mattjj:is-vjp-bit
4181fe49
Google-ML-Automation Update XLA dependency to use revision http://github.com/openxla/xla/c…
4f22f9d2
yashk2810 Canonicalize indexer.slice_sharding. Fixes https://github.com/jax-ml/…
7fbbe019
mattjj [hijax] improved dce interface
b1981cbb
Google-ML-Automation Merge pull request #35301 from mattjj:better-hijax-dce-interface
7f80f430
mattjj [hijax] fix followup on #35301
acad3005
Google-ML-Automation Merge pull request #35302 from mattjj:hijax-dce-fix
0cff9fe3
mattjj [hijax][remat3] start implementing remat on hijax
5f43a720
Google-ML-Automation Merge pull request #35304 from mattjj:remat-transform
53d6b350
mattjj [hijax][remat3] add basic remat_jaxpr
b17e41dd
Google-ML-Automation Merge pull request #35306 from mattjj:remat-jaxpr
dea31b6c
Google-ML-Automation Update XLA dependency to use revision http://github.com/openxla/xla/c…
8dac162f
superbobry [pyrefly] Fixed or suppressed type errors in Pallas backends
9ce7295f
Google-ML-Automation Merge pull request #35309 from superbobry:pyrefly
4d3395f0
Google-ML-Automation Update XLA dependency to use revision http://github.com/openxla/xla/c…
f47f0c20
chr1sj0nes [pallas:mgpu] Fix handling of `AbstractRef` in `_commute_transform`.
cd019717
Google-ML-Automation Merge pull request #33802 from jax-ml:dependabot/github_actions/actio…
a58b181f
phambinhfin Skip scaled matmul tests on CPU
11558844
jakevdp [test] fix deprecated mpmath import
b5c92015
superbobry More Pyrefly fixes in Pallas
0beb8372
Google-ML-Automation Merge pull request #35323 from jakevdp:mpmath-fix
adc7836d
Google-ML-Automation Merge pull request #35317 from superbobry:pyrefly
ad8c8c4f
jakevdp [typing] more pyrefly fixes
b1c5e118
superbobry Another batch of Pyrefly fixes
e673505c
apaszke [Mosaic TPU] Give more time for the changes to propagate to libtpu
db34816c
Arech8 Fix dangling std::string_view in keys of GetModuleImage() cache
01c30830
superbobry [jaxlib] Bundle _{tpu,triton,mosaic_gpu}_ext.pyi with jaxlib
22875809
danielsuo Pin mpmath to v1.3.0.
83471837
Google-ML-Automation Merge pull request #35332 from superbobry:pyrefly
b486aacf
superbobry Fixed `self.accum_ref` assertion in mosaic/pipeline.py
4dc41c30
danielsuo Roll back adding __eq__ and __hash__ to PyDevice.
920f6665
danielsuo [pmap] Replace `functools.lru_cache` with `util.cache` to fix stale `…
fc5ab2c1
danielsuo [lax_numpy] Adjust tolerance for np.float32 in trapezoid tests.
74b70ef1
dependabot[bot] Bump actions/setup-python from 6.1.0 to 6.2.0
ef24700b
dependabot[bot] Bump actions/cache from 4.3.0 to 5.0.3
9c33cd1c
Google-ML-Automation Merge pull request #35344 from jax-ml:dependabot/github_actions/actio…
45e29f6f
jakevdp [typing] suppress mypy redundant-cast
5d01b1fe
Google-ML-Automation Merge pull request #35347 from jax-ml:dependabot/github_actions/actio…
8ac9b74c
yashk2810 Re-enable some disabled tests in pjit_test.py
cc7365fa
Google-ML-Automation Merge pull request #35343 from jakevdp:fix-mypy-lint
677d7b0d
Google-ML-Automation Merge pull request #35331 from jakevdp:more-pyrefly-fixes
e12cf3d5
superbobry [mlir] Fixed the return type of `*{Attr,Type}.get` staticmethods
6dc44197
Google-ML-Automation Update XLA dependency to use revision http://github.com/openxla/xla/c…
d6d975c3
apaszke [Mosaic GPU] Add support for swizzle=16 in async SMEM->TMEM copies
b55efd81
apaszke Reverts 42c458cc318706bbbefbb45a752da6615a7123d3
776d3c38
superbobry More Pyrefly fixes
98f0ca20
apaszke [Pallas:MGPU] Allow SMEM->TMEM copies with unswizzled (but tiled) refs
1807a491
apaszke [Mosaic GPU] Relax N multiplicity restrictions for tcgen05 block-scal…
192601ce
apaszke [Mosaic GPU] Add support for fp4 tcgen05 MMAs that are both block-sca…
d037727f
Google-ML-Automation Merge pull request #35133 from mwhittaker:fault_tolerance_blog_tweaks
c139db7a
apaszke [Mosaic TPU] Expose the latest version from the TPU MosaicSerdePass
ab5ef253
belitskiy Add a light, blocking Mosaic B200 presubmit.
f0f74df5
apaszke [Pallas:MGPU] Add support for writes to WGMMA accumulator refs
fc9061d2
PatriosTheGreat [Mosaic:GPU] Temporary fix of ASAN test.
e16cf237
dependabot[bot] PR #35345: Bump actions/upload-artifact from 5.0.0 to 6.0.0
afa13c42
Google-ML-Automation Merge pull request #35367 from superbobry:pyrefly
012d6659
superbobry [mosaic] Verify that `tpu.memref_slice` always produces tile-aligned …
f59d8213
mattjj [hijax] mlir.lower_fun should have lo-jaxprs
b69d1c07
yueshengys [NFC] Remove some unnecessary skips to have better test coverage.
e76e398b
IvyZX Add Pallas SparseCore guide
6b89f7cb
Google-ML-Automation Merge pull request #35379 from mattjj:mlir-lower-fun-to-lojax
2cbf65c1
Google-ML-Automation Reverts f59d8213d42d124048de9ad93792bc17152134c7
7b56a865
yueshengys [Mosaic TPU] Add more complete support for 16-bit `cmpf` on TPU v5. P…
518b0d11
Google-ML-Automation Update XLA dependency to use revision http://github.com/openxla/xla/c…
0f31a1b0
brianwa84 Add SC support for pltpu.trace_value
2e3c8407
Google-ML-Automation Merge pull request #35334 from ROCm:ci_arech_fix_dangling_string_view
f6af222b
apaszke [Pallas:MGPU] Add a test for tcgen05 MMA that's both sparse and block…
90fdba0f
danielsuo [windows] Match triton importing in pallas_shape_poly_test to how it'…
eda5ec98
superbobry [jaxlib] Every MLIR dialect now comes with a .pyi
9169891b
danielsuo [lax_scipy] Adjust tolerance for np.float32 in trapezoid tests.
ab5f1971
danielsuo [lax_numpy] Skip ReducerWhere tests temporarily.
dc112482
apaszke [Mosaic TPU] Allow non-DMA semaphores in EnqueueDMA
ace7a790
Google-ML-Automation Merge pull request #35357 from superbobry:more-piys
1ae8be13
danielsuo [lax_numpy] Skip AdvancedIndexing tests temporarily.
122abbba
brianwa84 Enable some tests with TC-tiling with newer libtpu.
c6d2109c
brianwa84 Fix a bug in how input/output memory spaces are detected and propagat…
9c9dd77e
[typing] add type annotation for shard_map
709086fa
apaszke [Pallas:MGPU] Fix a race in dynamic_scheduling_loop
84a33c08
jakevdp [ci] Update pyrefly pre-commit to v0.54.0
a211a318
danielsuo [windows] Skip ops_test.test_binary_scalar if no TPU.
5c0ec3e8
danielsuo [lax_numpy] Skip setxor1 tests temporarily.
ac537c76
Google-ML-Automation Merge pull request #35404 from jakevdp:pyrefly-054
f5c39fed
Google-ML-Automation Merge pull request #35015 from IvyZX:scguide
100b05c8
jakevdp [pyrefly] remove ignore statements for issues fixed in v0.54
19e71a58
[typing] add type annotation for `jax.smap`.
b8171e35
Google-ML-Automation Merge pull request #35405 from jakevdp:pyrefly-ignore-statements
cda29303
Google-ML-Automation Merge pull request #34601 from ROCm:phambinh/ci_rocm_nn_tests
1f3139e0
Google-ML-Automation Merge pull request #35178 from ezhulenev:jax_sort_devices_by_process_…
015f1cbd
bythew3i [Pallas][Mosaic TPU] Add flag to disable semaphore checks per kernel.
77805318
jakevdp [typing] more pyrefly fixes in jax/_src/lax/
0461a3bb
Google-ML-Automation Merge pull request #35325 from jakevdp:pyrefly-fixes
742ad02a
superbobry Removed more dead code
480847f8
jakevdp [pyrefly] fix remaining jax/_src/lax errors
1e102aeb
danielsuo [windows] Skip triton export tests in pallas_shape_poly_test if trito…
7f250e0a
rdyro Explicitly check for disallowing TransformedRefs in higher-order JAX …
6a635c9a
levskaya [typing] pyrefly fixes in jax/_src/scipy/stats
b0d91345
levskaya [typing] pyrefly fixes in jax/_src/scipy/optimize
3b399a8f
superbobry [mosaic] Removed previously deprecated tpu.wait_dma
6869e28a
Google-ML-Automation Merge pull request #35414 from levskaya:pyrefly3
dc3f50ca
Google-ML-Automation Reverts 6869e28a4bde0f6d3a06a7c39858fe40cc44ff60
efa6bd9f
PatriosTheGreat [Mosaic:GPU] Simplify collective metadata initialization.
6f06ab80
Google-ML-Automation Update XLA dependency to use revision http://github.com/openxla/xla/c…
d0d6f426
akuegel Upgrade Abseil to LTS 20260107.1
ee4cea11
Google-ML-Automation Merge pull request #35412 from levskaya:pyrefly
4826da47
apaszke [Mosaic GPU] Add support for warp-level f8 MMAs
7192cf7d
apaszke [Pallas:TPU] Add tests for DMAs with regular semaphores
9e11f1c0
apaszke [Pallas:TPU] Add a test checking that dynamic shape exports require t…
f2017f78
alekstheod Set rpath based on the type (link_only)
af59a645
WindQAQ Add int1/uint1 dtypes.
c141c0fc
superbobry [mosaic] Another go at removing tpu.wait_dma
0b7d6672
apaszke [Mosaic GPU] Relax restrictions on N for sparse tcgen05 MMA
3f667fd4
yashk2810 Raise a better error if jit(shard_map) is bound with an AbstractMesh …
7af1783e
chr1sj0nes Fix hijax `VmapOf.batch_dim_rule` with `None` map dims.
6df7333b
mattjj [hijax] cleaner scan inc_rank/dec_rank api
4b87f031
hawkinsp [JAX] Bound tracebacks up to the enclosing `jit`.
6f2f0f36
hawkinsp Add a pre-commit hook that checks for Apache licence headers.
66b61418
brianwa84 Suppress still-broken test.
8f96afaa
danielsuo Roll back skipped tests that were failing on LLVM bit/byte-packing is…
873bebe1
danielsuo Update XLA dependency to use revision http://github.com/openxla/xla/c…
c3b53a26
danielsuo Add jaxlib version guard for updated make_c_api_client.
17cfbb81
danielsuo [jaxlib] Ported normalize_stubs.sh to Python
6428a91d
Google-ML-Automation Reverts 7af1783eb619f7e0236aa8c25fe1fc72881bb6b3
12344fc0
jakevdp Tie LinearizeTrace.tag to tangent_trace.tag
b6122922
Google-ML-Automation Merge pull request #35439 from mattjj:hijax-cleaner-scan-axis-spec
3b482638
Google-ML-Automation Merge pull request #35413 from jakevdp:pyrefly-lax-fixes
ed628980
jakevdp [hijax] lower jaxpr during discharge_state
fa369541
Google-ML-Automation [Pallas/TPU] Add tpu_info.get_tpu_info_for_chip
d237804a
danielsuo Skip pallas lowering determinism testOrderAgnostic test if not on TPU.
f82e2810
levskaya [typing] miscellaneous pyrefly typing fixes
85f6ef56
levskaya [typing] fixed remaining scipy pyrefly typing issues
47291b9f
naummo Generalize how op verifiers checking issuing core
008467a7
jakevdp Fix undefined name in ad_checkpoint
e81437f2
Google-ML-Automation Merge pull request #35219 from jakevdp:fix-tracer-tag
6d14b201
Google-ML-Automation Merge pull request #35452 from jakevdp:fix-ad-checkpoint
c08a1707
Google-ML-Automation Merge pull request #35448 from jakevdp:discharge-hijax
153937a3
danielsuo [linux-arm64] Reduce parallelism to 32 for nogil pytest cpu to avoid …
d873c960
Google-ML-Automation Merge pull request #35421 from levskaya:pyrefly4
019a87b1
jakevdp [pyrefly] fix errors in jax/interpreters/ad
d57e7c8b
danielsuo [pmap] Accelerate deprecation of PmapSharding.
11af98b4
Google-ML-Automation Merge pull request #35422 from levskaya:pyrefly5
e008d9d8
yashk2810 Add an experimental `top_level_all_gather` API (private for now).
a6df1f29
yashk2810 Add a test for replicated -> sharded+unreduced reshard
d506abdb
Google-ML-Automation Update XLA dependency to use revision http://github.com/openxla/xla/c…
08a51cea
Google-ML-Automation Merge pull request #35102 from ROCm:set_release_rpaths_to_rocm_so_tar…
717f99a4
apaszke [Pallas:MGPU] Allow inline_mgpu inside a warp context
df6415b4
Google-ML-Automation [Pallas][TPU kernel interpreter] Consistently use `pltu.HBM` to refer…
caf1324f
danielsuo Update XLA dependency to use revision http://github.com/openxla/xla/c…
0d37f593
danielsuo Skip pallas lowering determinism testOrderAgnostic test if not on TPU.
a573e865
danielsuo [linux-arm64] Reduce parallelism to 32 for nogil pytest cpu to avoid …
969da1b7
akuegel Reverts ee4cea11e08b92f23f464be2e7795f5eb9fe9557
143c2366
hawkinsp Remove references to tsl/platform/logging.h header.
89469861
superbobry [pallas:sc] Moved `jaxpr_call` into core Pallas
6400fa1a
joshuapjacob Fix typo in function parameter documentation
fd7c2aca
superbobry [mosaic] Added pattern canonicalizing tpu.memref_squeeze(memref.cast(…
3ad94844
cota [Mosaic GPU] fix annotation typo: s/pytype/pylint/
10a59a5e
apaszke [Mosaic TPU] Make sure to take the export flag into account when not …
b9926c39
superbobry [mosaic_gpu] Fixed mosaic-gpu-resolve-trivial-locations
e3adc6ad
PatriosTheGreat [Mosaic:GPU] Re-use the barrier buffers between executions to prevent…
1609c18f
PatriosTheGreat [Mosaic:GPU] Move collective kernel loading to the prepare stage.
f3bf01ad
PatriosTheGreat [Mosaic:GPU] Add test-case for several mosaic ops.
a25a24df
hawkinsp Reimplement weakref_lru_cache on top of a custom hash map type.
b2ccb32c
apaszke [Mosaic TPU] Make sure to allow unregistered dialects before deserial…
0a513d30
Google-ML-Automation Integrate LLVM at llvm/llvm-project@1053047a4be7
a30d6bc2
danielsuo Skip tests that are failing at libtpu=0.0.35, but passing at libtpu-0…
8350f362
danielsuo Bump jaxlib_extension_version to 412 due to LLVM integrate.
438e6946
danielsuo Skip tests that are failing at libtpu=0.0.35, but passing at libtpu-0…
0eb23146
tlongeri [Mosaic] Remove patterns to fold memref.cast into tpu.enqueue_dma and…
1899c328
ybaturina Rename `pypi_latest` libtpu-version-type to `pinned` and use pinned l…
abd765d1
charleshofer Add workflow for build wheels and running bazel tests on ROCm
bec3b790
yashk2810 Raise an error on `f32[4]{R:x} * f32[]`. For example `x: f32[4]{R:x} …
85ec15da
ybaturina Rename `pypi_latest` libtpu-version-type to `pinned` and use pinned l…
70a8149d
hawkinsp Reverts b2ccb32cabd9da9c99a4a9b7f163d4059ff448d5
41ae1eb6
naummo [Mosaic] Simplify GetCoreTypeOfParentOp
f154eb02
ybaturina Fix release workflow by using the folder with pre-downloaded jax whee…
cfe83cb2
ybaturina Fix release workflow by using the folder with pre-downloaded jax whee…
162b2ed0
yueshengys [Pallas][Mosaic TPU] More complete 1D block support in Pallas/Mosaic.
5c96b9ee
levskaya [pallas] rewrite pull_block_spec rule to work in terms of block_index…
0b565503
Google-ML-Automation Update XLA dependency to use revision http://github.com/openxla/xla/c…
b66a3c36
danielsuo Skip layout_test.test_layout_donation_mismatching_in_and_out_fails fo…
e5ab8054
danielsuo Skip layout_test.test_layout_donation_mismatching_in_and_out_fails fo…
7aa10a02
danielsuo Skip sparse random and dot general sampled ad tests.
500bbc20
danielsuo Skip sparse random and dot general sampled ad tests.
bb3362eb
danielsuo Prepare for JAX release 0.9.1
58cb6e55
Google-ML-Automation Update XLA dependency to use revision http://github.com/openxla/xla/c…
eba07b3e
Google-ML-Automation Update XLA dependency to use revision http://github.com/openxla/xla/c…
b0d504a3
Google-ML-Automation Add C addition to hopper matmul kernel.
d4204d58
danielsuo Merge remote-tracking branch 'origin/release/0.9.1' into postrelease/…
29ce624a
danielsuo Postrelease JAX v0.9.1.
5b9d90af
superbobry [jaxlib] :normalize_stubs now also fixes improper ``get`` classmethods
c5d55755
chr1sj0nes [mgpu] Minor cleanup in `try_cluster_cancel`.
95d95e3d
allanrenucci [Mosaic GPU][NFC] Remove duplicate layout inference rule for `vector.…
0cd1592e
Google-ML-Automation Merge pull request #35502 from superbobry:stubgen
b4e692ea
allanrenucci [Mosaic GPU][NFC] Use more precise type for `op` in `_optimization_ba…
e20119d1
superbobry [pallas:sc] Switched to using `pltpu.emit_pipeline` in the lowering
41430d32
vsytch [Mosaic] Add assembly format to tpu.delay_op
31fc0bd5
apaszke [Mosaic GPU] Fix for the latest LLVM version
44736b92
khasanovaa Add serialization support for Mosaic GPU kernels.
a3ffef5e
hawkinsp Regenerate type stubs for nanobind v2.12.0.
0ded2fed
allanrenucci [Mosaic GPU][NFC] Remove dead `constraints.BroadcastInDim`.
ed6d58f2
Google-ML-Automation Merge pull request #35501 from jax-ml:postrelease/0.9.1
edf82d10
hawkinsp Reverts 41ae1eb6ea50739feb91589705aa078b1495d40f
68b2565e
Google-ML-Automation Merge pull request #34702 from ROCm:add-run-bazel-test-rocm-rbe
ba54b343
Google-ML-Automation Merge pull request #35417 from joshuapjacob:patch-1
a192b434
apaszke [Mosaic GPU] Add support for untiled CP_ASYNC
a1b958ba
Google-ML-Automation Merge pull request #35453 from jakevdp:pyrefly-ad
4b423a65
charleshofer Remove nvidia_wheel_versions
2fe53184
charleshofer Make jaxlib targets visible
c7ec794c
charleshofer hipblas typedef fix
789df45d
charleshofer No GPU fail
d0e32075
mminutoli Wrap HIP inline functions in anonymous namespaces in vendor.h
caf56b60
dsicarov-amd SWDEV-512768 - Replace hipGetLastError with hipExtGetLastError
d8deb09f
charleshofer Add shared utility function get_rocm_version to test_util.py
6eee40dd
phambinhfin Fix hipSparse CSR algorithm mappings for ROCm 7
3128e0e7
phambinhfin Fix v_pages quantization and adjust test params for ROCm compatibilit…
cfddfa60
Arech8 Address LLVM assertion failure due to a multithreaded use. Update .gi…
cb904b38
Arech8 Add skip of test_is_finite() on Cuda (#565)
b78fe9a6
AratiGanesh Add rocm test requirements file (#570)
514fec32
charleshofer Let the unit tests use build.py for setting up Bazel commands for uni…
b64a5e40
gulsumgudukbay adding abort logic to rocm/jax (#590)
640e4a21
phambinhfin Skip is_finite tests on ROCm (not in Triton lowering for jax 0.8.0) (…
6eb4fdf5
phambinhfin Fix shared memory limit check for ROCm in test_dot (#596)
14b0125c
magaonka-amd Fix Numpy signatures test (#598)
92211a5d
Ruturaj4 fix merge arts
b202e1f4
gulsumgudukbay Enable RngShardingTests (#644)
4c60b2b0
mminutoli Enable test_variadic_reduce_window on ROCm (#647)
22ac32aa
magaonka-amd Skip sparse tests on ROCm due to hipSPARSE issue (#652)
8ecdd8b7
magaonka-amd Update sparse test skip messages in v0.8.2 (#653)
cc543662
magaonka-amd Skip sparse tests on ROCm due to hipSPARSE issue (#652)
1b6c9e72
magaonka-amd Update sparse test skip messages in v0.8.2 (#653)
f0e90395
magaonka-amd Skip sparse tests on ROCm due to hipSPARSE issue (#652)
0cc3b988
magaonka-amd Update sparse test skip messages in v0.8.2 (#653)
47321964
AratiGanesh Enable testMultivariateNormalSingularCovariance on ROCm (#666)
e5714928
AratiGanesh Skip test_tridiagonal_solve on ROCm due to hipSPARSE numerical errors…
84ad452e
gulsumgudukbay Update Skip Reason Outputs (#663)
3c295779
magaonka-amd Skip sparse tests on ROCm due to hipSPARSE issue (#652)
7648e11b
magaonka-amd Update sparse test skip messages in v0.8.2 (#653)
e0d2d1e9
magaonka-amd Skip testCudaArrayInterfaceOnNonCudaFails on ROCm platform (#677)
a186d625
magaonka-amd Skip sparse tests on ROCm due to hipSPARSE issue (#652)
9983dd00
magaonka-amd Update sparse test skip messages in v0.8.2 (#653)
c45c49a9
magaonka-amd Skip sparse tests on ROCm due to hipSPARSE issue (#652)
14d1b171
magaonka-amd Update sparse test skip messages in v0.8.2 (#653)
ec394330
AratiGanesh Add ROCm encoding for test_struct_encoding_determinism (#683)
d18c7571
magaonka-amd Remove 'mean' from unsupported params for jnp.var (#689)
717cd20f
magaonka-amd Implement approx_tanh for ROCm using OCML tanh function (#691)
02ae6c5b
AratiGanesh Skipping testEighTinyNorm due to hipSolver issues (#697)
f2f9dd18
gulsumgudukbay Abort detection CI workflow (#688)
6957e42d
gulsumgudukbay Abort-Detection: Fix halt-for-connection input (#712)
7d684aa5
alekstheod alekstheod requested a review 97 days ago
alekstheod alekstheod closed this 97 days ago

Login to write a write a comment.

Login via GitHub

Reviewers
No reviews
Assignees
No one assigned
Labels
Milestone