Commits

Branch selector

 main

User selector

beckerhe

Datepicker

All time

Commit History

Commits on Oct 30, 2025

• Add serialization for ffi::Attribute

  Show description for dd65fcc
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  dd65fcc

  Copy full SHA for dd65fcc

Commits on Oct 29, 2025

• Add proto serialization for GpuComputeCapability

  Show description for 327d6db
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  327d6db

  Copy full SHA for 327d6db

Commits on Oct 28, 2025

• Add proto serialization for RocmComputeCapability

  Show description for 7239f89
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  7239f89

  Copy full SHA for 7239f89

• Add missing dependency to TSL target

  Show description for 3bf96fa
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  3bf96fa

  Copy full SHA for 3bf96fa

Commits on Oct 27, 2025

• Move Attribute types from call_frame.cc into attribute_map.cc

  Show description for 5abb0d9
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  5abb0d9

  Copy full SHA for 5abb0d9

Commits on Oct 14, 2025

• Replace `CustomCallThunk::Slice` by `ShapedSlice`

  Show description for ba157a8
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  ba157a8

  Copy full SHA for ba157a8

Commits on Oct 13, 2025

• Move custom call handler resolution into CustomCallThunk

  Show description for 2f6ec2d
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  2f6ec2d

  Copy full SHA for 2f6ec2d

• Remove unused nvtx_utils target

  Show description for 4099327
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  4099327

  Copy full SHA for 4099327

• Remove MLIR-related unused code from CollectiveThunk

  Show description for 02f3eb1
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  02f3eb1

  Copy full SHA for 02f3eb1

Commits on Oct 10, 2025

• Sort dumped execution files before checking.

  Show description for 0af33e8
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  0af33e8

  Copy full SHA for 0af33e8

Commits on Oct 9, 2025

• Introduce dumping of thunk metadata to file

  Show description for a8f3a79
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  a8f3a79

  Copy full SHA for a8f3a79

Commits on Oct 8, 2025

• Introduce TemporaryDirectory type

  Show description for 521a6a9
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  521a6a9

  Copy full SHA for 521a6a9

Commits on Oct 6, 2025

• Make ptx_kernel_test work on P100

  Show description for c2182c2
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  c2182c2

  Copy full SHA for c2182c2

• Pin Docker images in XLA presubmit benchmarks

  Show description for 87608c0
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  87608c0

  Copy full SHA for 87608c0

Commits on Oct 2, 2025

• Add serialization support for CuBlasLtMatmulThunk

  Show description for 860578b
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  860578b

  Copy full SHA for 860578b

Commits on Oct 1, 2025

• Make TMA test kernel work on B200 as well

  Show description for c4f0ac1
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  c4f0ac1

  Copy full SHA for c4f0ac1

Commits on Sep 30, 2025

• Remove `if_google` conditions for `google/protobuf` dependencies in XLA BUILD files.

  Show description for 6ae1506
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  6ae1506

  Copy full SHA for 6ae1506

Commits on Sep 29, 2025

• Add missing header dependencies to cudnn_frontend target

  Show description for 8282534
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  8282534

  Copy full SHA for 8282534

• Remove `@com_google_protobuf//:protobuf` from XLA default dependencies.

  Show description for 6b56032
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  6b56032

  Copy full SHA for 6b56032

• Protobuf dependency cleanup

  Show description for a9a1213
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  a9a1213

  Copy full SHA for a9a1213

Commits on Sep 25, 2025

• Fix indent handling in SequentialThunk::ToString

  Show description for d669004
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  d669004

  Copy full SHA for d669004

Commits on Sep 24, 2025

• Introduce ThunkId

  Show description for 6a78abc
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  6a78abc

  Copy full SHA for 6a78abc

Commits on Sep 18, 2025

• Fix GPU AOT Compilation test

  Show description for 6757b75
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  6757b75

  Copy full SHA for 6757b75

Commits on Sep 16, 2025

• Fix FP8 data type handling for cublasLt gemms

  Show description for 788eb17
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  788eb17

  Copy full SHA for 788eb17

Commits on Sep 10, 2025

• Fix fission handling for cublasLt

  Show description for cb77a12
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  cb77a12

  Copy full SHA for cb77a12

Commits on Sep 9, 2025

• Make TSAN work with non instrumented code (CUDA libraries)

  Show description for a3f807d
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  a3f807d

  Copy full SHA for a3f807d

Commits on Sep 4, 2025

• Skip FP8 fused attention tests on cuDNN 9.10.0.

  Show description for 5816cef
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  5816cef

  Copy full SHA for 5816cef

Commits on Sep 3, 2025

• Provide the module name via a dedicated getter in xla::Executable

  Show description for e30e2d3
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  e30e2d3

  Copy full SHA for e30e2d3

• Remove LeakCheckDisabler on CUDA 13+

  Show description for 967ca81
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  967ca81

  Copy full SHA for 967ca81

Commits on Sep 1, 2025

• Make the autotuning cache key depend on the cudnn version

  Show description for d9a75de
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  d9a75de

  Copy full SHA for d9a75de

• Remove HloGraphDumper dependency from xla::Executable

  Show description for df559c1
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  df559c1

  Copy full SHA for df559c1

• Reverts 234b3471efb8182838adc0e10d8a0f2179309f3d

  Show description for 9ccf9dd
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  9ccf9dd

  Copy full SHA for 9ccf9dd

• Exclude feature extensions from compute capability string in TF's GPU device details.

  Show description for 0082221
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  0082221

  Copy full SHA for 0082221

Commits on Aug 29, 2025

• Remove HLO and Autotuner dependency from CublasLtMatmulThunk

  Show description for a8b50a0
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  a8b50a0

  Copy full SHA for a8b50a0

Commits on Aug 28, 2025

• Reverts 87b9835bc735769fcc613709dab1acf6f31026e8

  Show description for 234b347
  

  beckerhe

  authored and

  Google-ML-Automation

  committed
  234b347

  Copy full SHA for 234b347

Pagination

Previous

Next