Commits

Breadcrumbs

History for

onmain

User selector

All users

Datepicker

All time

Commit History

Commits on Apr 29, 2025

• [pt2d] Add reorder_comms_preserving_peak_memory pass (#146562)

  Show description for a4a7716
  

  wconstab

  authored and

  pytorchmergebot

  committed
  a4a7716

  Copy full SHA for a4a7716

Commits on Apr 28, 2025

• [DTensor] Error on illegal view op during sharding prop (#149764)

  Show description for 0c52ee1
  

  wconstab

  authored and

  pytorchmergebot

  committed
  0c52ee1

  Copy full SHA for 0c52ee1

• [DTensor] make test_dtensor_ops report dtensor_args (#152045)

  Show description for efeed72
  

  wconstab

  authored and

  pytorchmergebot

  committed
  efeed72

  Copy full SHA for efeed72

Commits on Apr 27, 2025

• Fix broken URLs (#152237)

  Show description for e2f9759
  

  shoumikhin

  authored and

  pytorchmergebot

  committed
  e2f9759

  Copy full SHA for e2f9759

Commits on Apr 25, 2025

• [CP] Use TorchFunctionMode to dispatch SDPA for CP (#147902)

  Show description for 6aa9280
  

  fegin

  authored and

  pytorchmergebot

  committed
  6aa9280

  Copy full SHA for 6aa9280

• Fix xrefs (#151888)

  Show description for 9e50c21
  

  shoumikhin

  authored and

  pytorchmergebot

  committed
  9e50c21

  Copy full SHA for 9e50c21

Commits on Apr 22, 2025

• logging start of torch elastic workers. (#150849)

  Show description for a7ccd96
  

  aschhabra

  authored and

  pytorchmergebot

  committed
  a7ccd96

  Copy full SHA for a7ccd96

• Fix DTensorTestBase to barrier with device ids (#150896)

  Show description for 3380a46
  

  wanchaol

  authored and

  pytorchmergebot

  committed
  3380a46

  Copy full SHA for 3380a46

• [c10d][fr] Fix another bug when we should continue when the op list is empty (#151798)

  Show description for 95abc0f
  

  fduwjj

  authored and

  pytorchmergebot

  committed
  95abc0f

  Copy full SHA for 95abc0f

Commits on Apr 19, 2025

• Add api to enable/disable NaN detector per-PG (#151723)

  Show description for 2673ea4
  

  wconstab

  authored and

  pytorchmergebot

  committed
  2673ea4

  Copy full SHA for 2673ea4

Commits on Apr 18, 2025

• c10d/Store: add nonblocking mode to queue_pop (#151485)

  Show description for 98c8927
  

  d4l3k

  authored and

  pytorchmergebot

  committed
  98c8927

  Copy full SHA for 98c8927

Commits on Apr 17, 2025

• [C10D] avoid computing global_rank when group_rank is used (#151373)

  Show description for 9e235c5
  

  wconstab

  authored and

  pytorchmergebot

  committed
  9e235c5

  Copy full SHA for 9e235c5

• [c10d][fr] Fix script for uneven reduce scatter and update test cases (#151475)

  Show description for 6f9ffaa
  

  fduwjj

  authored and

  pytorchmergebot

  committed
  6f9ffaa

  Copy full SHA for 6f9ffaa

Commits on Apr 16, 2025

• improve noop elimination for view (#151095)

  Show description for 2fd2692
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  2fd2692

  Copy full SHA for 2fd2692

• [dynamo] unimplemented -> unimplemented_v2 in variables/builtin.py (#151145)

  Show description for 1f29190
  

  williamwen42

  authored and

  pytorchmergebot

  committed
  1f29190

  Copy full SHA for 1f29190

• [inductor][comms] skip reorder_for_locality for wait nodes (#150074)

  Show description for c7400d0
  

  xmfan

  authored and

  pytorchmergebot

  committed
  c7400d0

  Copy full SHA for c7400d0

Commits on Apr 15, 2025

• [c10d][fr] Record each individual collective being coalesced (#151238)

  Show description for 71e7dcd
  

  fduwjj

  authored and

  pytorchmergebot

  committed
  71e7dcd

  Copy full SHA for 71e7dcd

• [c10d][fr] Enable FR analysis script for rest of all coalesce op (#151247)

  Show description for ae648f0
  

  fduwjj

  authored and

  pytorchmergebot

  committed
  ae648f0

  Copy full SHA for ae648f0

• Add @requires_multicast_support to test_multimem_all_gather (#151227)

  Show description for 7e5f6dc
  

  pragupta

  authored and

  pytorchmergebot

  committed
  7e5f6dc

  Copy full SHA for 7e5f6dc

• [c10d][fr] Enable FR analysis script for all fast-path coalesce op (#151243)

  Show description for 48b4bc1
  

  fduwjj

  authored and

  pytorchmergebot

  committed
  48b4bc1

  Copy full SHA for 48b4bc1

• test_store: fix timeout for test_queues (#151252)

  Show description for 4ede670
  

  d4l3k

  authored and

  pytorchmergebot

  committed
  4ede670

  Copy full SHA for 4ede670

Commits on Apr 14, 2025

• [ROCm][CI] Enable distributed CI on MI300 (#150667)

  Show description for 001695c
  

  jithunnair-amd

  authored and

  pytorchmergebot

  committed
  001695c

  Copy full SHA for 001695c

Commits on Apr 13, 2025

• [DTensor] Add DTensor redistribute fwd/bwd datatype conversion to enable SimpleFSDP mixed precision training (#150740)

  Show description for 1c5619e
  

  ruisizhang123

  authored and

  pytorchmergebot

  committed
  1c5619e

  Copy full SHA for 1c5619e

Commits on Apr 12, 2025

• [dtensor] add op support for torch._grouped_mm (#151072)

  Show description for 7dd2ed1
  

  tianyu-l

  authored and

  pytorchmergebot

  committed
  7dd2ed1

  Copy full SHA for 7dd2ed1

Commits on Apr 11, 2025

• [DTensor] Fix empty shard global-offset calculation (#150862)

  Show description for c3bc6b3
  

  wconstab

  authored and

  pytorchmergebot

  committed
  c3bc6b3

  Copy full SHA for c3bc6b3

• c10d/Store: add queues (#150969)

  Show description for bb60e82
  

  d4l3k

  authored and

  pytorchmergebot

  committed
  bb60e82

  Copy full SHA for bb60e82

• [c10d][fr] Add logging of nccl_version into fr and its dump (#151048)

  Show description for 6b32255
  

  fduwjj

  authored and

  pytorchmergebot

  committed
  6b32255

  Copy full SHA for 6b32255

• [dtensor] add op support for torch.cumsum (#151071)

  Show description for d385179
  

  tianyu-l

  authored and

  pytorchmergebot

  committed
  d385179

  Copy full SHA for d385179

• c10d/Store: add clone feature (#150966) (#150966) (#151045)

  Show description for 8b5e717
  

  d4l3k

  authored and

  pytorchmergebot

  committed
  8b5e717

  Copy full SHA for 8b5e717

• [c10d][fr] Fix the false positive in the dtype check in fr analysis script (#151063)

  Show description for 48132de
  

  fduwjj

  authored and

  pytorchmergebot

  committed
  48132de

  Copy full SHA for 48132de

• Reapply "ProcessGroupGloo: support lazy_init (#150801)" (#151031)

  Show description for df4e529
  

  d4l3k

  authored and

  pytorchmergebot

  committed
  df4e529

  Copy full SHA for df4e529

Commits on Apr 10, 2025

• Revert "c10d/Store: add clone feature (#150966)"

  Show description for abe41c5
  pytorchmergebot

  committed
  abe41c5

  Copy full SHA for abe41c5

• [async TP] Fix handling of case where scatter dim = 0 for 2D output tensor (#150935)

  Show description for e7ed50f
  

  danielvegamyhre

  authored and

  pytorchmergebot

  committed
  e7ed50f

  Copy full SHA for e7ed50f

• Revert "ProcessGroupGloo: support lazy_init (#150801)"

  Show description for 73f3d6d
  pytorchmergebot

  committed
  73f3d6d

  Copy full SHA for 73f3d6d

• [elastic][test] fix race condition in test_barrier_timeout_rank_tracing (#150768)

  Show description for 1a1a32c
  

  cdzhan

  authored and

  pytorchmergebot

  committed
  1a1a32c

  Copy full SHA for 1a1a32c

Pagination

Previous

Next