Commits

Branch selector

 master

User selector

All users

All time

Commit History

Commits on Apr 21, 2026

• Automated Code Change

  Show description for 7ec45b4
  tensorflower-gardener

  committed
  7ec45b4

  Copy full SHA for 7ec45b4

• Automated Code Change

  Show description for 309af3a
  tensorflower-gardener

  committed
  309af3a

  Copy full SHA for 309af3a

• Automated Code Change

  Show description for 68add99
  tensorflower-gardener

  committed
  68add99

  Copy full SHA for 68add99

• Remove redundant A - A => 0 rule

  Show description for 3085cd8
  

  dsharletg

  authored and

  tensorflower-gardener

  committed
  3085cd8

  Copy full SHA for 3085cd8

• Add axis count check in Mean op.

  Show description for 886bfdc
  

  lynne0326

  authored and

  tensorflower-gardener

  committed
  886bfdc

  Copy full SHA for 886bfdc

• [IFRT IR] Keep delete_options and device_list for IFRT IR programs

  Show description for e7eb58e
  

  ICGog

  authored and

  tensorflower-gardener

  committed
  e7eb58e

  Copy full SHA for e7eb58e

• [XLA] Improve conditional code motion for broadcasts to avoid performance and memory issues.

  Show description for 3594e60
  

  blakehechtman

  authored and

  tensorflower-gardener

  committed
  3594e60

  Copy full SHA for 3594e60

• Make LLVMCompiler hooks private and thread-safe

  Show description for 5125ae1
  

  ermilovmaxim

  authored and

  tensorflower-gardener

  committed
  5125ae1

  Copy full SHA for 5125ae1

Commits on Apr 20, 2026

• PR #41176: [ROCm] Fix gpu_compiler_test by relaxing an fp8 support check

  Show description for 5c5e74f
  

  Eetusjo

  authored and

  tensorflower-gardener

  committed
  5c5e74f

  Copy full SHA for 5c5e74f

• [XLA:CPU] Add regression test for tensordot OOM

  Show description for 5302ef2
  

  seantalts

  authored and

  tensorflower-gardener

  committed
  5302ef2

  Copy full SHA for 5302ef2

• Extend IfrtRestoreVariable lowering to handle variable copies.

  Show description for a8b20f9
  

  xiaoxlu

  authored and

  tensorflower-gardener

  committed
  a8b20f9

  Copy full SHA for a8b20f9

• [XLA][Partitioner][Bug-Fix] Fix empty axis_refs crash in MeshAxesReplicaGroupList.

  Show description for 43ccf69
  

  Varcho

  authored and

  tensorflower-gardener

  committed
  43ccf69

  Copy full SHA for 43ccf69

• Support XLA metadata propagation for a common TF compilation code path

  Show description for d976e27
  tensorflower-gardener

  committed
  d976e27

  Copy full SHA for d976e27

• Update hermetic Python references to @rules_ml_toolchain

  Show description for d647618
  

  yuriivcs

  authored and

  tensorflower-gardener

  committed
  d647618

  Copy full SHA for d647618

• [XLA] Support instructions with multiple operands in HostOffloader.

  Show description for 66d42fa
  tensorflower-gardener

  committed
  66d42fa

  Copy full SHA for 66d42fa

• create SingleDeviceSharding when possible

  Show description for 1222cfb
  tensorflower-gardener

  committed
  1222cfb

  Copy full SHA for 1222cfb

• Move jobs to run on self-hosted CPU runner.

  Show description for d27fec7
  

  jialei-chen

  authored and

  tensorflower-gardener

  committed
  d27fec7

  Copy full SHA for d27fec7

• Fix a dangling pointer bug in FindAndWrapOffloadedComputations.

  Show description for 87976e4
  

  patrick-toulme

  authored and

  tensorflower-gardener

  committed
  87976e4

  Copy full SHA for 87976e4

• Roll forward with a fix

  Show description for 9973560
  

  junwhanahn

  authored and

  tensorflower-gardener

  committed
  9973560

  Copy full SHA for 9973560

• Fixed memory allocation in while_op_test

  Show description for e61413a
  

  dmiltr3

  authored and

  tensorflower-gardener

  committed
  e61413a

  Copy full SHA for e61413a

• [QwixOdmlQuantizationBoundary] Fix MLIR assertion failure and metadata leakage in U-Net quantization.

  Show description for 060dd46
  

  jshin1394

  authored and

  tensorflower-gardener

  committed
  060dd46

  Copy full SHA for 060dd46

• Optimize MultiOutputFusion reachability updates.

  Show description for a659740
  tensorflower-gardener

  committed
  a659740

  Copy full SHA for a659740

• Add z-scoring benchmark to norm_benchmark_test

  Show description for 0fc46d4
  

  dsharletg

  authored and

  tensorflower-gardener

  committed
  0fc46d4

  Copy full SHA for 0fc46d4

• [XLA:GPU] Print tiled parameters correctly.

  Show description for 30e5b10
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  30e5b10

  Copy full SHA for 30e5b10

• Use sharded queue in slinky_threadpool

  Show description for 9820056
  

  alexander-shaposhnikov

  authored and

  tensorflower-gardener

  committed
  9820056

  Copy full SHA for 9820056

• [XLA:GPU] emit reduce

  Show description for c76767e
  

  metaflow

  authored and

  tensorflower-gardener

  committed
  c76767e

  Copy full SHA for c76767e

• Make failure log more informative

  Show description for 3e2aef8
  

  abhigunj

  authored and

  tensorflower-gardener

  committed
  3e2aef8

  Copy full SHA for 3e2aef8

• [XLA:GPU] Fix CSE during tiling.

  Show description for 7aeeef9
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  7aeeef9

  Copy full SHA for 7aeeef9

• fix sorting.hlo test

  Show description for 94c476a
  

  ermilovmaxim

  authored and

  tensorflower-gardener

  committed
  94c476a

  Copy full SHA for 94c476a

• Rollback of PR #40920

  Show description for 25d431c
  

  mkuperst

  authored and

  tensorflower-gardener

  committed
  25d431c

  Copy full SHA for 25d431c

• [XLA:GPU] Emit bitcast with the same logic as tiling, i.e., transpose, reshape, trt.

  Show description for 7fd8f68
  tensorflower-gardener

  committed
  7fd8f68

  Copy full SHA for 7fd8f68

• [XLA] Treat computations with a different number of parameters as different.

  Show description for 37072e6
  

  dimitar-asenov

  authored and

  tensorflower-gardener

  committed
  37072e6

  Copy full SHA for 37072e6

• [XLA:CPU] Add FAST_COMPILE tests to CI using test_cpu_fast_compile argument.

  Show description for b383e6b
  

  seantalts

  authored and

  tensorflower-gardener

  committed
  b383e6b

  Copy full SHA for b383e6b

• Remove AffineMap serialization/printing from indexing_map_serialization

  Show description for 95a1412
  tensorflower-gardener

  committed
  95a1412

  Copy full SHA for 95a1412

• Move constants to the right when canonicalize a chain of products

  Show description for 43eaa0c
  tensorflower-gardener

  committed
  43eaa0c

  Copy full SHA for 43eaa0c