Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Aug 15, 2024

• [XLA] Enable passing of precomputed analyses when evaluating the result of an HLO instruction via HloEvaluator.

  Show description for 85d3abf
  

  tensorflower-gardener

  authored and

  copybara-github

  committed
  85d3abf

  Copy full SHA for 85d3abf

• Update nanobind to v2.1.0 in XLA.

  Show description for 76520d7
  

  hawkinsp

  authored and

  copybara-github

  committed
  76520d7

  Copy full SHA for 76520d7

• Make ExecutorCache deal only in ordinals rather than the StreamExecutorConfig which only contains an ordinal now.

  Show description for dc36530
  

  klucke

  authored and

  copybara-github

  committed
  dc36530

  Copy full SHA for dc36530

• [XLA:SPMD] Remove `LookaheadUserSharding` in sharding propagation.

  Show description for 8689542
  

  ZixuanJiang

  authored and

  copybara-github

  committed
  8689542

  Copy full SHA for 8689542

• Stop running dot_algorithm_support_test on p100.

  Show description for 1b42023
  

  klucke

  authored and

  copybara-github

  committed
  1b42023

  Copy full SHA for 1b42023

• Integrate LLVM at llvm/llvm-project@1115dee248e6

  Show description for 8af2cd8
  

  tensorflower-gardener

  authored and

  copybara-github

  committed
  8af2cd8

  Copy full SHA for 8af2cd8

• PR #16014: Gloo support for MacOS: Second try

  Show description for 1b01e0b
  

  heiner

  authored and

  copybara-github

  committed
  1b01e0b

  Copy full SHA for 1b01e0b

• [XLA:GatherScatter] fix gather/scatter shape inference

  Show description for 4517a0c
  

  tomnatan30

  authored and

  copybara-github

  committed
  4517a0c

  Copy full SHA for 4517a0c

• [XLA] [NFC] Start runnning functional_hlo_runner_test continuously

  Show description for 3af08a7
  

  cheshire

  authored and

  copybara-github

  committed
  3af08a7

  Copy full SHA for 3af08a7

• [XLA] [NFC] Remove dead line

  Show description for dc80b01
  

  cheshire

  authored and

  copybara-github

  committed
  dc80b01

  Copy full SHA for dc80b01

• [XLA:GPU] Add BlockLevelParameters cache.

  Show description for 4281ae2
  

  olegshyshkov

  authored and

  copybara-github

  committed
  4281ae2

  Copy full SHA for 4281ae2

• [XLA:GPU] Implement fusing int4 parameters into Triton dots.

  Show description for 4cfd0f8
  

  loislo

  authored and

  copybara-github

  committed
  4cfd0f8

  Copy full SHA for 4cfd0f8

• PR #16098: Revert "PR #15989: [GPU] Upgrade cuDNN frontend to 1.6.0."

  Show description for 5afb50f
  

  sergachev

  authored and

  copybara-github

  committed
  5afb50f

  Copy full SHA for 5afb50f

• PR #16102: Lower threshold for F64 in //xla/service/gpu/model:hlo_op_profiler_test

  Show description for efa74cd
  

  sergey-kozub

  authored and

  copybara-github

  committed
  efa74cd

  Copy full SHA for efa74cd

• Automated Code Change

  Show description for 252c977
  

  tensorflower-gardener

  authored and

  copybara-github

  committed
  252c977

  Copy full SHA for 252c977

• [XLA] More random dynamic copy test cases

  Show description for 6450b14
  

  vsytch

  authored and

  copybara-github

  committed
  6450b14

  Copy full SHA for 6450b14

• [XLA] Don't forget to program while loop backpointers

  Show description for ea466d3
  

  vsytch

  authored and

  copybara-github

  committed
  ea466d3

  Copy full SHA for ea466d3

• Integrate LLVM at llvm/llvm-project@6de04e6fe8b1

  Show description for 6779ea1
  

  tensorflower-gardener

  authored and

  copybara-github

  committed
  6779ea1

  Copy full SHA for 6779ea1

• Add more CUDA and CUDNN JSON links.

  Show description for f9c92b1
  

  tensorflower-gardener

  authored and

  copybara-github

  committed
  f9c92b1

  Copy full SHA for f9c92b1

• Remove gpu_stream from StreamExecutorConfig. All users have been moved to the stream_executor::FindStream API instead.

  Show description for 3978dde
  

  klucke

  authored and

  copybara-github

  committed
  3978dde

  Copy full SHA for 3978dde

• Allow type-level specification of exhaustive binary 16-bit parameter packing

  Show description for f800d6f
  

  Gregory Pataky

  authored and

  copybara-github

  committed
  f800d6f

  Copy full SHA for f800d6f

Commits on Aug 14, 2024

• Support TSL compilation on Windows with CUDA support.

  Show description for 7ae6696
  

  eaplatanios

  authored and

  copybara-github

  committed
  7ae6696

  Copy full SHA for 7ae6696

• [xla:cpu] NFC: Move XLA:CPU runtime to xla/backends/cpu component

  Show description for 5b5f580
  

  ezhulenev

  authored and

  copybara-github

  committed
  5b5f580

  Copy full SHA for 5b5f580

• Add easier debugging for exhaustive tests

  Show description for ac359c4
  

  Gregory Pataky

  authored and

  copybara-github

  committed
  ac359c4

  Copy full SHA for ac359c4

• Build target visibility change

  Show description for 2adb9b3
  

  tensorflower-gardener

  authored and

  copybara-github

  committed
  2adb9b3

  Copy full SHA for 2adb9b3

• Tighten tolerances for Sub exhaustive_binary_16_bit_test

  Show description for a570a42
  

  Gregory Pataky

  authored and

  copybara-github

  committed
  a570a42

  Copy full SHA for a570a42

• [easy] [XLA][HostOffloading] Separate utils used in Host Offloader

  Show description for aa23400
  

  tensorflower-gardener

  authored and

  copybara-github

  committed
  aa23400

  Copy full SHA for aa23400

• Improve exhaustive test range logging

  Show description for 2e8be6d
  

  Gregory Pataky

  authored and

  copybara-github

  committed
  2e8be6d

  Copy full SHA for 2e8be6d

• Add OpenXLA banner to XLA README.md

  Show description for 9161219
  

  GleasonK

  authored and

  copybara-github

  committed
  9161219

  Copy full SHA for 9161219

• Introduce hermetic CUDA in Google ML projects.

  Show description for cb1541c
  

  tensorflower-gardener

  authored and

  copybara-github

  committed
  cb1541c

  Copy full SHA for cb1541c

• Add a version of CreateBuffersForAsyncHostToDevice that takes a custom layout.

  Show description for 3c40cdd
  

  tensorflower-gardener

  authored and

  copybara-github

  committed
  3c40cdd

  Copy full SHA for 3c40cdd

• Fix XLA build for windows when built in JAX context

  Show description for c768961
  

  vam-google

  authored and

  copybara-github

  committed
  c768961

  Copy full SHA for c768961

• [XLA] Updates UsesBeforeValueDefinition to consider all types of async calls

  Show description for 2de7712
  

  tensorflower-gardener

  authored and

  copybara-github

  committed
  2de7712

  Copy full SHA for 2de7712

• [GPU] Re-Enable sharding of autotuning by default.

  Show description for bff259c
  

  tensorflower-gardener

  authored and

  copybara-github

  committed
  bff259c

  Copy full SHA for bff259c

• Integrate Triton up to [803c5883](https://github.com/openai/triton/commits/803c5883046b920b31404b10fafce547e5590265)

  Show description for cb304cf
  

  vwbaker

  authored and

  copybara-github

  committed
  cb304cf

  Copy full SHA for cb304cf

Pagination

Previous

Next