Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on May 22, 2025

• [XLA:GPU][NFC] Add a helper to emit HLO-style broadcast.

  Show description for b1a4d9a
  

  pifon2a

  authored and

  Google-ML-Automation

  committed
  b1a4d9a

  Copy full SHA for b1a4d9a

• Defines an alternative implementation of strategy shaving for IOPDDL problems.

  Show description for a095c8c
  Google-ML-Automation

  committed
  a095c8c

  Copy full SHA for a095c8c

• [XLA:CPU] Remove unused StoreOp

  Show description for c5b6e16
  

  WillFroom

  authored and

  Google-ML-Automation

  committed
  c5b6e16

  Copy full SHA for c5b6e16

• [XLA:GPU]: Fix test for oneshot all-reduce

  Show description for ddf7348
  

  sohaibiftikhar

  authored and

  Google-ML-Automation

  committed
  ddf7348

  Copy full SHA for ddf7348

• #sdy Add unique channel id to created `stablehlo.all_reduce`.

  Show description for f040466
  

  tomnatan30

  authored and

  Google-ML-Automation

  committed
  f040466

  Copy full SHA for f040466

• [XLA][graphviz] Visualize nested thunks as clusters

  Show description for ee64cb1
  

  basioli-k

  authored and

  Google-ML-Automation

  committed
  ee64cb1

  Copy full SHA for ee64cb1

• [XLA:CPU] Only create the kernel api at the lowering stage

  Show description for 31754c8
  

  WillFroom

  authored and

  Google-ML-Automation

  committed
  31754c8

  Copy full SHA for 31754c8

• PR #26762: [ROCm] adjust hipblaslt workspace size for gfx942

  Show description for 98acf9d
  

  ScXfjiang

  authored and

  Google-ML-Automation

  committed
  98acf9d

  Copy full SHA for 98acf9d

• PR #26828: [ROCm] Filter out empty amdgpu target

  Show description for 67d9c78
  

  hsharsha

  authored and

  Google-ML-Automation

  committed
  67d9c78

  Copy full SHA for 67d9c78

• [XLA:CPU] Reuse LLVM struct types when possible during lowering

  Show description for 63c556d
  

  WillFroom

  authored and

  Google-ML-Automation

  committed
  63c556d

  Copy full SHA for 63c556d

• [XLA:GPU] Fix symbolic tiling for nested fusions with point dimensions.

  Show description for a3f1700
  

  chsigg

  authored and

  Google-ML-Automation

  committed
  a3f1700

  Copy full SHA for a3f1700

• [XLA:GPU] Get kernel arity information from KernelTrait type.

  Show description for 7f1349f
  

  olegshyshkov

  authored and

  Google-ML-Automation

  committed
  7f1349f

  Copy full SHA for 7f1349f

• [XLA:CPU][xnnpack] Xnnpack backend for generating executables for autotuning

  Show description for 286ea67
  

  basioli-k

  authored and

  Google-ML-Automation

  committed
  286ea67

  Copy full SHA for 286ea67

• [XLA:GPU] Implement ApplyConfig for CustomKernel autotuner backend.

  Show description for 5b66909
  

  derdrdirk

  authored and

  Google-ML-Automation

  committed
  5b66909

  Copy full SHA for 5b66909

• Automated Code Change

  Show description for 6485624
  Google-ML-Automation

  committed
  6485624

  Copy full SHA for 6485624

• [XLA:GPU] Hoist bitcasts past transposes in nest gemm fusion pass.

  Show description for 3740294
  

  chsigg

  authored and

  Google-ML-Automation

  committed
  3740294

  Copy full SHA for 3740294

• Block the transfer to the device until it's done.

  Show description for 5c7c520
  Google-ML-Automation

  committed
  5c7c520

  Copy full SHA for 5c7c520

• Automated Code Change

  Show description for d500acf
  Google-ML-Automation

  committed
  d500acf

  Copy full SHA for d500acf

• Automated Code Change

  Show description for a768e0e
  Google-ML-Automation

  committed
  a768e0e

  Copy full SHA for a768e0e

• Automated Code Change

  Show description for 5ff1a7c
  Google-ML-Automation

  committed
  5ff1a7c

  Copy full SHA for 5ff1a7c

• Set remaining profile options and perform validation on the advanced_configuration keys.

  Show description for ae6c5c4
  

  sannidhyachauhan

  authored and

  Google-ML-Automation

  committed
  ae6c5c4

  Copy full SHA for ae6c5c4

• Automated Code Change

  Show description for 460a92c
  Google-ML-Automation

  committed
  460a92c

  Copy full SHA for 460a92c

• Expose `getAxisSizes` API for `HloSharding` such that we can call it directly in Python with JAX.

  Show description for fbcf6ad
  

  ZixuanJiang

  authored and

  Google-ML-Automation

  committed
  fbcf6ad

  Copy full SHA for fbcf6ad

• [GPU] Add CopyThunk::FromProto

  Show description for 681fdbb
  

  apivovarov

  authored and

  Google-ML-Automation

  committed
  681fdbb

  Copy full SHA for 681fdbb

• [XLA] Add support for sharing computations to XlaBuilder, and use it in MHLO to HLO conversion.

  Show description for ef9fdc4
  

  mkuperst

  authored and

  Google-ML-Automation

  committed
  ef9fdc4

  Copy full SHA for ef9fdc4

• [XLA][Numerics][HLO Value Tracking] Adds a unit test for AddOriginalTest pass

  Show description for 1b04510
  

  jcai19

  authored and

  Google-ML-Automation

  committed
  1b04510

  Copy full SHA for 1b04510

• Reverts fd23612846c9720713fde47c85df743ca45ac240

  Show description for 1a2a360
  Google-ML-Automation

  committed
  1a2a360

  Copy full SHA for 1a2a360

• [xla:gpu] Cleanup GpuClique comparison operators

  Show description for 698304e
  

  ezhulenev

  authored and

  Google-ML-Automation

  committed
  698304e

  Copy full SHA for 698304e

• Add VLOGs for transfers

  Show description for 38914dd
  

  hhb

  authored and

  Google-ML-Automation

  committed
  38914dd

  Copy full SHA for 38914dd

• [PJRT:SE] Add VLOGs to record the start and finish of a device execution.

  Show description for d864098
  Google-ML-Automation

  committed
  d864098

  Copy full SHA for d864098

Commits on May 21, 2025

• Use shape layout (if present) in `CommonPjRtClient::CreateUninitializedBuffer`

  Show description for 2adf2ef
  

  Gregory Pataky

  authored and

  Google-ML-Automation

  committed
  2adf2ef

  Copy full SHA for 2adf2ef

• Drop shard_count for a bunch more tests to align with the actual test count.

  Show description for 12a9a0e
  

  pizzud

  authored and

  Google-ML-Automation

  committed
  12a9a0e

  Copy full SHA for 12a9a0e

• [xla] Add more details to rendezvous CHECKs

  Show description for f2061c6
  

  ezhulenev

  authored and

  Google-ML-Automation

  committed
  f2061c6

  Copy full SHA for f2061c6

• Replace `HostCallBack` in `TransferRawDataToSubBuffer` with `BlockHostUntilDone`

  Show description for 64314b7
  

  hhb

  authored and

  Google-ML-Automation

  committed
  64314b7

  Copy full SHA for 64314b7

• [XLA] Handle cloning scheduled async ops

  Show description for ad3f6fc
  

  vsytch

  authored and

  Google-ML-Automation

  committed
  ad3f6fc

  Copy full SHA for ad3f6fc

Pagination

Previous

Next