Commits

Branch selector

 main

User selector

BoyuanFeng

Datepicker

All time

Commit History

Commits on Nov 5, 2024

• [AOTAutograd] Handle edge cases for donated buffer & enable in oss (#139669)

  Show description for 87059d4
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  87059d4

  Copy full SHA for 87059d4

Commits on Oct 29, 2024

• [Flex Attention] Paged Attention (#137164)

  Show description for 68134a3
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  68134a3

  Copy full SHA for 68134a3

Commits on Sep 28, 2024

• [Flex Attention] fix block size order (#136657)

  Show description for 2822432
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  2822432

  Copy full SHA for 2822432

Commits on Sep 27, 2024

• [Flex Attention] fix block size order (#136657)

  Show description for b42f1e3
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  b42f1e3

  Copy full SHA for b42f1e3

Commits on Sep 25, 2024

• [Flex Attention] fix block size order (#136657)

  Show description for 529b6ab
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  529b6ab

  Copy full SHA for 529b6ab

Commits on Sep 10, 2024

• [FlexAttention] Add broadcast support for kv batch dimension (#135505)

  Show description for 6e13f5e
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  6e13f5e

  Copy full SHA for 6e13f5e

Commits on Sep 4, 2024

• [Flex Attention] update __getitem__ without tree_map_only to support compile (#134627)

  Show description for 09a339f
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  09a339f

  Copy full SHA for 09a339f

Commits on Aug 7, 2024

• [CUDAGraph] Warn once if too many distinct sizes (#132832)

  Show description for 6a348e5
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  6a348e5

  Copy full SHA for 6a348e5

Commits on Jul 26, 2024

• [AOT Autograd] Donated Buffer (#130580)

  Show description for 40cc5c0
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  40cc5c0

  Copy full SHA for 40cc5c0

• [CUDAGraph] skip cudagraph if too many distinct sizes (#131387)

  Show description for aebfd3d
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  aebfd3d

  Copy full SHA for aebfd3d

• [CUDAGraph] Type annotation for cudagraph_trees.py (#131621)

  Show description for 16d7cb5
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  16d7cb5

  Copy full SHA for 16d7cb5

• [CUDAGraph] Type annotation for cudagraph_trees.py (#131621)

  Show description for 16699c7
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  16699c7

  Copy full SHA for 16699c7

Commits on Jul 18, 2024

• [CUDAGraph] Add operator.mul to skip list for find_input_mutations (#130986)

  Show description for 8bf0be7
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  8bf0be7

  Copy full SHA for 8bf0be7

Commits on Jul 17, 2024

• [ts-migration] Support RaiseException, prim::Unitialized, prim::Enter, and prim::Exit (#129416)

  Show description for 90105a4
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  90105a4

  Copy full SHA for 90105a4

• [ts-migration] Support RaiseException, prim::Unitialized, prim::Enter, and prim::Exit (#129416)

  Show description for f0faecd
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  f0faecd

  Copy full SHA for f0faecd

• [ts-migration] Support RaiseException, prim::Unitialized, prim::Enter, and prim::Exit (#129416)

  Show description for ef05112
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  ef05112

  Copy full SHA for ef05112

Commits on Jul 14, 2024

• [cudagraph] fallback to eager if re-record too many times (#129349)

  Show description for fa5f572
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  fa5f572

  Copy full SHA for fa5f572

Commits on Jun 29, 2024

• [ts-migration] support prim::SetAttr and fix prim::GetAttr (#129440)

  Show description for 3ef44df
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  3ef44df

  Copy full SHA for 3ef44df

Commits on Jun 27, 2024

• [CUDAGraph Trees] Enable input mutation support in OSS (#129184)

  Show description for 0b7e8df
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  0b7e8df

  Copy full SHA for 0b7e8df

Commits on Jun 18, 2024

• [ts migration] Support prim::tolist and aten::len (#128894)

  Show description for 04a5d32
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  04a5d32

  Copy full SHA for 04a5d32

• [CUDAGraph] add more docs for cudagraph trees (#127963)

  Show description for 4399871
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  4399871

  Copy full SHA for 4399871

Commits on Jun 4, 2024

• [ts-migration] support aten::__is__, aten::__isnot__, aten::__not__, profiler::_record_function_enter_new, profiler::_record_function_exit (#127656)

  Show description for 2ad0e41
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  2ad0e41

  Copy full SHA for 2ad0e41

Commits on Jun 2, 2024

• [ts migration] support aten::dim, aten::len, aten::__getitem__ (#127593)

  Show description for 2cef2fc
  

  31 peoplecommitted
  2cef2fc

  Copy full SHA for 2cef2fc

Commits on May 30, 2024

• [cudagraph] improve log for mutating static input tensor addresses (#127145)

  Show description for ff23c5b
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  ff23c5b

  Copy full SHA for ff23c5b

• [torchscript] Handle prim::device and prim::dtype (#127466)

  Show description for 4afc5c7
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  4afc5c7

  Copy full SHA for 4afc5c7

Commits on May 29, 2024

• [Cudagraph] better support for streams (#126809)

  Show description for 2b72e2a
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  2b72e2a

  Copy full SHA for 2b72e2a

Commits on Apr 24, 2024

• [cudagraph] add config for cudagraph managed input mutation support (#124754)

  Show description for b91f83f
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  b91f83f

  Copy full SHA for b91f83f

Commits on Apr 19, 2024

• [CUDAGraphTree] Support mutated inputs from prior cudagraph pool (#123231)

  Show description for 9a71d12
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  9a71d12

  Copy full SHA for 9a71d12

Commits on Apr 18, 2024

• [Export] Add runtime assert to non-strict export (#123681)

  Show description for aa2da0c
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  aa2da0c

  Copy full SHA for aa2da0c

Commits on Apr 2, 2024

• Add inline constraints to non-strict exported program (#123017)

  Show description for 64d7430
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  64d7430

  Copy full SHA for 64d7430

Commits on Mar 28, 2024

• Remove several `expectedFailureNonStrict` (#122802)

  Show description for 574a8cc
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  574a8cc

  Copy full SHA for 574a8cc

Commits on Mar 27, 2024

• Add non strict inline constraints and runtime assertions to non-strict exported program (#122722)

  Show description for b693fff
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  b693fff

  Copy full SHA for b693fff

Commits on Mar 20, 2024

• Fix `_chunk_cat.out` issue (#122076)

  Show description for 2f6e8e8
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  2f6e8e8

  Copy full SHA for 2f6e8e8

Commits on Mar 13, 2024

• CUDA fast path for `_chunk_cat()` (#120678)

  Show description for aaff8d2
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  aaff8d2

  Copy full SHA for aaff8d2

• Support `call_method` in DDPOptimizer (#121771)

  Show description for 0c1ac44
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  0c1ac44

  Copy full SHA for 0c1ac44

Pagination

Previous

Next