Commits

Branch selector

 main

User selector

mlazos

Datepicker

All time

Commit History

Commits on Dec 17, 2024

• [foreach-map] Add tests for backward (#143282)

  Show description for cb4c614
  

  mlazos

  authored and

  pytorchmergebot

  committed
  cb4c614

  Copy full SHA for cb4c614

Commits on Dec 12, 2024

• [Dynamo] only import einops if version is lower than 0.7.0 (#142847)

  Show description for 357e261
  

  mlazos

  authored and

  pytorchmergebot

  committed
  357e261

  Copy full SHA for 357e261

• [Dynamo] Add missing tensor builtins to allowed functions (#142841)

  Show description for 9701c50
  

  mlazos

  authored and

  pytorchmergebot

  committed
  9701c50

  Copy full SHA for 9701c50

Commits on Dec 11, 2024

• [foreach_map] Initial foreach map HOP impl for inference (#142098)

  Show description for de313f1
  

  mlazos

  authored and

  pytorchmergebot

  committed
  de313f1

  Copy full SHA for de313f1

• [Dynamo] Add register_hook as in-graph tensor method (#142820)

  Show description for 539c46b
  

  mlazos

  authored and

  pytorchmergebot

  committed
  539c46b

  Copy full SHA for 539c46b

• [Dynamo] Refactor to use install subgraph method in higher order ops (#141384)

  Show description for 082124a
  

  mlazos

  authored and

  pytorchmergebot

  committed
  082124a

  Copy full SHA for 082124a

• [Dynamo] Initial deduplication pass impl (#141383)

  Show description for c31543c
  

  mlazos

  authored and

  pytorchmergebot

  committed
  c31543c

  Copy full SHA for c31543c

• [Dynamo] add debug logging for graph region expansion (#141382)

  Show description for 49e4307
  

  mlazos

  authored and

  pytorchmergebot

  committed
  49e4307

  Copy full SHA for 49e4307

• [Dynamo] Implement graph region tracking for deduplication (#141381)

  Show description for 96c36a6
  

  mlazos

  authored and

  pytorchmergebot

  committed
  96c36a6

  Copy full SHA for 96c36a6

Commits on Dec 10, 2024

• Add support for bfloat16 atomic adds in fbcode (#141857)

  Show description for a3abe1a
  

  mlazos

  authored and

  pytorchmergebot

  committed
  a3abe1a

  Copy full SHA for a3abe1a

Commits on Nov 15, 2024

• Ensure index for state guard construction is a source (#140515)

  Show description for ffd5197
  

  mlazos

  authored and

  pytorchmergebot

  committed
  ffd5197

  Copy full SHA for ffd5197

• Support tensor betas in Adam and AdamW (#134171)

  Show description for 1fd4757
  

  mlazos

  authored and

  pytorchmergebot

  committed
  1fd4757

  Copy full SHA for 1fd4757

Commits on Nov 9, 2024

• Make size a property with an assertion (#139794)

  Show description for 8b2e385
  

  mlazos

  authored and

  pytorchmergebot

  committed
  8b2e385

  Copy full SHA for 8b2e385

Commits on Nov 8, 2024

• [Dynamo] allow dynamic callables on tensor variables (#137940)

  Show description for ea0f60e
  

  mlazos

  authored and

  pytorchmergebot

  committed
  ea0f60e

  Copy full SHA for ea0f60e

Commits on Nov 6, 2024

• [Dynamo] Support tensor mro without source (#139838)

  Show description for d622b49
  

  mlazos

  authored and

  pytorchmergebot

  committed
  d622b49

  Copy full SHA for d622b49

Commits on Oct 28, 2024

• [Dynamo] Refactor wrap_fx_proxy (#138933)

  Show description for 46132dc
  

  mlazos

  authored and

  pytorchmergebot

  committed
  46132dc

  Copy full SHA for 46132dc

Commits on Oct 19, 2024

• [Dynamo] Disable torch function compilation during guard execution and in compiled bytecode (#137669)

  Show description for a20a17f
  

  mlazos

  authored and

  pytorchmergebot

  committed
  a20a17f

  Copy full SHA for a20a17f

Commits on Oct 17, 2024

• Support more foreach ops for tensor beta support (#134170)

  Show description for 0b2c12c
  

  mlazos

  authored and

  pytorchmergebot

  committed
  0b2c12c

  Copy full SHA for 0b2c12c

Commits on Oct 15, 2024

• [Dynamo] Disable torch function compilation during guard execution and in compiled bytecode (#137669)

  Show description for bf0b670
  

  mlazos

  authored and

  pytorchmergebot

  committed
  bf0b670

  Copy full SHA for bf0b670

• Fix autograd function calls without context arg (#137809)

  Show description for 8316f9b
  

  mlazos

  authored and

  pytorchmergebot

  committed
  8316f9b

  Copy full SHA for 8316f9b

Commits on Oct 10, 2024

• Add warning for non static grads in optimizer variable (#137554)

  Show description for d50d5df
  

  mlazos

  authored and

  pytorchmergebot

  committed
  d50d5df

  Copy full SHA for d50d5df

Commits on Oct 9, 2024

• [Dynamo] add flex attention mode test (#137121)

  Show description for 5f1759a
  

  mlazos

  authored and

  pytorchmergebot

  committed
  5f1759a

  Copy full SHA for 5f1759a

• [Dynamo] Handle torch function subclass/mode dispatch on generic tensor methods (#137119)

  Show description for d5785d4
  

  mlazos

  authored and

  pytorchmergebot

  committed
  d5785d4

  Copy full SHA for d5785d4

• [Dynamo] Handle extracted unbound tensor methods (#137227)

  Show description for 0a304d9
  

  mlazos

  authored and

  pytorchmergebot

  committed
  0a304d9

  Copy full SHA for 0a304d9

• [Dynamo] Move flex attention torch function mode to traceable HOP file (#137120)

  Show description for b3f30c9
  

  mlazos

  authored and

  pytorchmergebot

  committed
  b3f30c9

  Copy full SHA for b3f30c9

• [Dynamo] Ensure torch function modes are dispatched on builtin ops (#137117)

  Show description for 27dee93
  

  mlazos

  authored and

  pytorchmergebot

  committed
  27dee93

  Copy full SHA for 27dee93

• [Dynamo] Remove ignored modes from torch function mode stack guard (#135503) (#137116)

  Show description for 38afac2
  

  mlazos

  authored and

  pytorchmergebot

  committed
  38afac2

  Copy full SHA for 38afac2

• [Dynamo] Remove ignored modes workaround (#135502) (#137115)

  Show description for 108b469
  

  mlazos

  authored and

  pytorchmergebot

  committed
  108b469

  Copy full SHA for 108b469

• [Dynamo] Trace enter/exit of TorchFunctionModes (#135422) (#137114)

  Show description for e41dffb
  

  mlazos

  authored and

  pytorchmergebot

  committed
  e41dffb

  Copy full SHA for e41dffb

Commits on Oct 8, 2024

• Update torchbind expecttest from landrace (#137453)

  Show description for 8ca15e8
  

  mlazos

  authored and

  pytorchmergebot

  committed
  8ca15e8

  Copy full SHA for 8ca15e8

Commits on Oct 7, 2024

• [Dynamo] add flex attention mode test (#137121)

  Show description for 144665d
  

  mlazos

  authored and

  pytorchmergebot

  committed
  144665d

  Copy full SHA for 144665d

• [Dynamo] Handle torch function subclass/mode dispatch on generic tensor methods (#137119)

  Show description for d255b34
  

  mlazos

  authored and

  pytorchmergebot

  committed
  d255b34

  Copy full SHA for d255b34

• [Dynamo] Handle extracted unbound tensor methods (#137227)

  Show description for 14eabd6
  

  mlazos

  authored and

  pytorchmergebot

  committed
  14eabd6

  Copy full SHA for 14eabd6

• [Dynamo] Move flex attention torch function mode to traceable HOP file (#137120)

  Show description for 68151fd
  

  mlazos

  authored and

  pytorchmergebot

  committed
  68151fd

  Copy full SHA for 68151fd

• [Dynamo] Ensure torch function modes are dispatched on builtin ops (#137117)

  Show description for 941be41
  

  mlazos

  authored and

  pytorchmergebot

  committed
  941be41

  Copy full SHA for 941be41

Pagination

Previous

Next