Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Nov 22, 2024

• Add reshard to all_gather pass

  Show description for 7fa60d9
  

  jordiae

  authored and

  copybara-github

  committed
  7fa60d9

  Copy full SHA for 7fa60d9

• Integrate LLVM at llvm/llvm-project@556ea5265a25

  Show description for ceaa0c9
  

  metaflow

  authored and

  copybara-github

  committed
  ceaa0c9

  Copy full SHA for ceaa0c9

• Integrate LLVM at llvm/llvm-project@a12e79a85fc1

  Show description for ca0deec
  

  metaflow

  authored and

  copybara-github

  committed
  ca0deec

  Copy full SHA for ca0deec

• Set implicitTrunc on APInt creation

  Show description for 07df09c
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  07df09c

  Copy full SHA for 07df09c

Commits on Nov 21, 2024

• Remove patch that is not needed anymore.

  Show description for d5411ea
  

  akuegel

  authored and

  copybara-github

  committed
  d5411ea

  Copy full SHA for d5411ea

• Integrate LLVM at llvm/llvm-project@33fcd6acc755

  Show description for 52beb13
  

  metaflow

  authored and

  copybara-github

  committed
  52beb13

  Copy full SHA for 52beb13

Commits on Nov 20, 2024

• Op documentation update.

  Show description for 263eb04
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  263eb04

  Copy full SHA for 263eb04

• Add sdy.all_gather

  Show description for 3cdd029
  

  jordiae

  authored and

  copybara-github

  committed
  3cdd029

  Copy full SHA for 3cdd029

• Consider all prefixes of a tensor sharding so that largest available prefix can be assigned to factors.

  Show description for ddc15aa
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  ddc15aa

  Copy full SHA for ddc15aa

• Count factor-axes candidates accurately for when one axes is prefix of another, and extend from existing assignment.

  Show description for f9220c3
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  f9220c3

  Copy full SHA for f9220c3

Commits on Nov 19, 2024

• Integrate LLVM at llvm/llvm-project@b03a747fc0fc

  Show description for bb66029
  

  metaflow

  authored and

  copybara-github

  committed
  bb66029

  Copy full SHA for bb66029

• Integrate LLVM at llvm/llvm-project@6e1acdcdc1b3

  Show description for 40a9fd7
  

  metaflow

  authored and

  copybara-github

  committed
  40a9fd7

  Copy full SHA for 40a9fd7

Commits on Nov 18, 2024

• [SDY] fix JAX layouts tests for Shardy.

  Show description for f75bd45
  

  Varcho

  authored and

  copybara-github

  committed
  f75bd45

  Copy full SHA for f75bd45

• [StableHLO] Add VhloDialect to the dependent dialects of vhlo-to-version

  Show description for 0747261
  

  GleasonK

  authored and

  copybara-github

  committed
  0747261

  Copy full SHA for 0747261

• Use FactorAxesPair instead of FactorAxesAssignmentCandidate to hold the best one.

  Show description for 6454408
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  6454408

  Copy full SHA for 6454408

• Integrate LLVM at llvm/llvm-project@64c455077abe

  Show description for 8a7d79e
  

  metaflow

  authored and

  copybara-github

  committed
  8a7d79e

  Copy full SHA for 8a7d79e

• Introduce and use AxesWithTail struct to represent axes with a tailing AxisRefAttr.

  Show description for af77414
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  af77414

  Copy full SHA for af77414

• Update tie-break so that larger axes based on AxesWithTail comparator has a preference.

  Show description for cc98c69
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  cc98c69

  Copy full SHA for cc98c69

• Factor out finding factor-axes pairs and their counts.

  Show description for 33fcea5
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  33fcea5

  Copy full SHA for 33fcea5

Commits on Nov 16, 2024

• [StableHLO] Don't require inlining for shape refinement.

  Show description for 740f65a
  

  GleasonK

  authored and

  copybara-github

  committed
  740f65a

  Copy full SHA for 740f65a

Commits on Nov 15, 2024

• Integrate LLVM at llvm/llvm-project@b3134fa23383

  Show description for 30a0925
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  30a0925

  Copy full SHA for 30a0925

• Add a tailing AxisRefAttr to FactorAxesPair, so that the FactorAxesPair can represent any axes, excluding its most-minor axis, that is a prefix of an existing factor sharding.

  Show description for b76f0cd
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  b76f0cd

  Copy full SHA for b76f0cd

• Refactor to use FactorAxesAssignmentCandidate.assignTo api for assigning the picked axes to the picked factor.

  Show description for d0be1a1
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  d0be1a1

  Copy full SHA for d0be1a1

• Use FactorAxesAssignmentCandidate struct to simplify the logic that picks the best factor axes pair.

  Show description for 52b9dae
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  52b9dae

  Copy full SHA for 52b9dae

• Use DenseMap of factor-axis pairs instead of a vector whose elements correspond to factors.

  Show description for f5edbfa
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  f5edbfa

  Copy full SHA for f5edbfa

• Find common axes deterministically by not depending on DenseMap iteration order to find the best factor-axes pair.

  Show description for 9234712
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  9234712

  Copy full SHA for 9234712

• Op documentation update.

  Show description for f7847c5
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  f7847c5

  Copy full SHA for f7847c5

Commits on Nov 14, 2024

• Integrate LLVM at llvm/llvm-project@03730cdd3d10

  Show description for 5e34ee5
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  5e34ee5

  Copy full SHA for 5e34ee5

• [SDY:nit] update summary of sharding group op to be more descriptive.

  Show description for c3f3c8e
  

  Varcho

  authored and

  copybara-github

  committed
  c3f3c8e

  Copy full SHA for c3f3c8e

• Change sharding rule of @PartialReduce custom call so we don't propagate sideways between the operands or results along the reduction dim.

  Show description for 4e1ad0d
  

  tomnatan30

  authored and

  copybara-github

  committed
  4e1ad0d

  Copy full SHA for 4e1ad0d

• Integrate LLVM at llvm/llvm-project@97298853b4de

  Show description for 3f43aab
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  3f43aab

  Copy full SHA for 3f43aab

• Propagate through broadcasts at the 2nd iteration of op priority propagation.

  Show description for afaaa93
  

  tomnatan30

  authored and

  copybara-github

  committed
  afaaa93

  Copy full SHA for afaaa93

• Op documentation update.

  Show description for f02c30d
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  f02c30d

  Copy full SHA for f02c30d

• Fix typos in docs.

  Show description for 879e94b
  

  tomnatan30

  authored and

  copybara-github

  committed
  879e94b

  Copy full SHA for 879e94b

• Refactor to use a std::pair to hold factor index and axis refs.

  Show description for 7549f48
  

  Google-ML-Automation

  authored and

  copybara-github

  committed
  7549f48

  Copy full SHA for 7549f48

Pagination

Previous

Next