Commits

Commit history

1. Commits on Sep 1, 2025

   • [CUDA] Reuse blocks with record_stream during CUDA Graph capture in the CUDACachingAllocator (#158352)

     Show description for 190c391
     

     2 peopleauthored and

     pytorchmergebot

     committed
     190c391View commit details

     Copy full SHA for 190c391

     View code at this pointBrowse repository at this point
2. Commits on Aug 29, 2025

   • [CI] Add basic CUDA 13.0 periodic test (#161013)

     Show description for 303f514
     

     3 peopleauthored and

     pytorchmergebot

     committed
     303f514View commit details

     Copy full SHA for 303f514

     View code at this pointBrowse repository at this point
3. Commits on Aug 27, 2025

   • Revert "Refactor CUDAAllocatorConfig to reuse AcceleratorAllocatorConfig (#150312)" (#161628)

     Show description for b9c6aa1
     

     guangyey

     authored and

     pytorchmergebot

     committed
     b9c6aa1View commit details

     Copy full SHA for b9c6aa1

     View code at this pointBrowse repository at this point

   • Revert "Deprecate overleap functions in CUDAAllocatorConfig, use AcceleratorAllocatorConfig instead (#156165)" (#161627)

     Show description for b7b9fb9
     

     guangyey

     authored and

     pytorchmergebot

     committed
     b7b9fb9View commit details

     Copy full SHA for b7b9fb9

     View code at this pointBrowse repository at this point

   • Revert "Generalize torch._C._set_allocator_settings to be generic (#156175)" (#161626)

     Show description for c03d8d4
     

     guangyey

     authored and

     pytorchmergebot

     committed
     c03d8d4View commit details

     Copy full SHA for c03d8d4

     View code at this pointBrowse repository at this point

   • Revert "Back out "Deprecate overleap functions in CUDAAllocatorConfig, use AcceleratorAllocatorConfig instead (#156165)" (#160999)" (#161625)

     Show description for 06ddaf1
     

     guangyey

     authored and

     pytorchmergebot

     committed
     06ddaf1View commit details

     Copy full SHA for 06ddaf1

     View code at this pointBrowse repository at this point

   • Revert "Back out "Refactor CUDAAllocatorConfig to reuse AcceleratorAllocatorConfig (#150312)" (#161002)"

     Show description for 69c7b16
     pytorchmergebot

     committed
     69c7b16View commit details

     Copy full SHA for 69c7b16

     View code at this pointBrowse repository at this point
4. Commits on Aug 26, 2025

   • Back out "Refactor CUDAAllocatorConfig to reuse AcceleratorAllocatorConfig (#150312)" (#161002)

     Show description for a03cc53
     

     joshuuuasu

     authored and

     pytorchmergebot

     committed
     a03cc53View commit details

     Copy full SHA for a03cc53

     View code at this pointBrowse repository at this point
5. Commits on Aug 25, 2025

   • [nccl symm mem] don't use arg for mempool, correctly use symmetric registration in hooks (#161238)

     Show description for 726dce3
     

     ngimel

     authored and

     pytorchmergebot

     committed
     726dce3View commit details

     Copy full SHA for 726dce3

     View code at this pointBrowse repository at this point
6. Commits on Aug 22, 2025

   • Refactoring TensorImpl by using constexpr and std::is_same_v (#161043)

     Show description for 2beffb3
     

     fffrog

     authored and

     pytorchmergebot

     committed
     2beffb3View commit details

     Copy full SHA for 2beffb3

     View code at this pointBrowse repository at this point
7. Commits on Aug 20, 2025

   • Back out "Deprecate overleap functions in CUDAAllocatorConfig, use AcceleratorAllocatorConfig instead (#156165)" (#160999)

     Show description for a818fa7
     

     joshuuuasu

     authored and

     pytorchmergebot

     committed
     a818fa7View commit details

     Copy full SHA for a818fa7

     View code at this pointBrowse repository at this point
8. Commits on Aug 19, 2025

   • fabric detection - fix build on an old toolkit (#160984)

     Show description for 0d19541
     

     ngimel

     authored and

     pytorchmergebot

     committed
     0d19541View commit details

     Copy full SHA for 0d19541

     View code at this pointBrowse repository at this point
9. Commits on Aug 18, 2025

   • harden fabric checks for symmetric memory (#160790)

     Show description for 0254646
     

     ngimel

     authored and

     pytorchmergebot

     committed
     0254646View commit details

     Copy full SHA for 0254646

     View code at this pointBrowse repository at this point

   • Revert "Remove guard_size_oblivious from default contiguity python check, and add aten.sym_is_contiguous. (#159197)"

     Show description for b82aa3d
     pytorchmergebot

     committed
     b82aa3dView commit details

     Copy full SHA for b82aa3d

     View code at this pointBrowse repository at this point
10. Commits on Aug 16, 2025

    • Remove guard_size_oblivious from default contiguity python check, and add aten.sym_is_contiguous. (#159197)

      Show description for e444cd2
      

      laithsakka

      authored and

      pytorchmergebot

      committed
      e444cd2View commit details

      Copy full SHA for e444cd2

      View code at this pointBrowse repository at this point
11. Commits on Aug 13, 2025

    • Replace TORCH_INTERNAL_ASSERT with TORCH_CHECK (#160411)

      Show description for 8eee08d
      

      fffrog

      authored and

      pytorchmergebot

      committed
      8eee08dView commit details

      Copy full SHA for 8eee08d

      View code at this pointBrowse repository at this point
12. Commits on Aug 12, 2025

    • fix retaining multimem in symmetric memory (#160343)

      Show description for be53f60
      

      ngimel

      authored and

      pytorchmergebot

      committed
      be53f60View commit details

      Copy full SHA for be53f60

      View code at this pointBrowse repository at this point
13. Commits on Aug 11, 2025

    • [codemod] Fix unreachable-break issue in caffe2/c10/cuda/CUDAFunctions.cpp +2 (#160257)

      Show description for 68a4b4b
      

      r-barnes

      authored and

      pytorchmergebot

      committed
      68a4b4bView commit details

      Copy full SHA for 68a4b4b

      View code at this pointBrowse repository at this point

    • [CUDA] Remove the uncessary CUDA_GUARD (#160249)

      Show description for dc0d18e
      

      fffrog

      authored and

      pytorchmergebot

      committed
      dc0d18eView commit details

      Copy full SHA for dc0d18e

      View code at this pointBrowse repository at this point
14. Commits on Aug 8, 2025

    • [MPS] Extend `index_put` to complex types (#160159)

      Show description for 28ccc9e
      

      malfet

      authored and

      pytorchmergebot

      committed
      28ccc9eView commit details

      Copy full SHA for 28ccc9e

      View code at this pointBrowse repository at this point

    • Add DeviceAllocator as the base device allocator (#138222)

      Show description for d7114f0
      

      guangyey

      authored and

      pytorchmergebot

      committed
      d7114f0View commit details

      Copy full SHA for d7114f0

      View code at this pointBrowse repository at this point

    • [MPS] coalesce for sparse tensors (#159729)

      Show description for 7f4cb4a
      

      2 peopleauthored and

      pytorchmergebot

      committed
      7f4cb4aView commit details

      Copy full SHA for 7f4cb4a

      View code at this pointBrowse repository at this point
15. Commits on Aug 7, 2025

    • Revert "Add DeviceAllocator as the base device allocator (#138222)"

      Show description for f3a4d74
      pytorchmergebot

      committed
      f3a4d74View commit details

      Copy full SHA for f3a4d74

      View code at this pointBrowse repository at this point
16. Commits on Aug 6, 2025

    • [Reland] Migrate ScalarType to headeronly (#159911)

      Show description for 1690c0c
      

      janeyx99

      authored and

      pytorchmergebot

      committed
      1690c0cView commit details

      Copy full SHA for 1690c0c

      View code at this pointBrowse repository at this point

    • Add DeviceAllocator as the base device allocator (#138222)

      Show description for f7a66da
      

      guangyey

      authored and

      pytorchmergebot

      committed
      f7a66daView commit details

      Copy full SHA for f7a66da

      View code at this pointBrowse repository at this point
17. Commits on Aug 5, 2025

    • [C10] Add `Scalar::isUnsigned()` method (#159877)

      Show description for 882d50c
      

      malfet

      authored and

      pytorchmergebot

      committed
      882d50cView commit details

      Copy full SHA for 882d50c

      View code at this pointBrowse repository at this point

    • Generalize torch._C._set_allocator_settings to be generic (#156175)

      Show description for 908c5cc
      

      guangyey

      authored and

      pytorchmergebot

      committed
      908c5ccView commit details

      Copy full SHA for 908c5cc

      View code at this pointBrowse repository at this point

    • Deprecate overleap functions in CUDAAllocatorConfig, use AcceleratorAllocatorConfig instead (#156165)

      Show description for c114585
      

      guangyey

      authored and

      pytorchmergebot

      committed
      c114585View commit details

      Copy full SHA for c114585

      View code at this pointBrowse repository at this point

    • Refactor CUDAAllocatorConfig to reuse AcceleratorAllocatorConfig (#150312)

      Show description for ae1a706
      

      guangyey

      authored and

      pytorchmergebot

      committed
      ae1a706View commit details

      Copy full SHA for ae1a706

      View code at this pointBrowse repository at this point

    • Fix AllocatorConfig potential SIO issue (#159629)

      Show description for 56d19a5
      

      guangyey

      authored and

      pytorchmergebot

      committed
      56d19a5View commit details

      Copy full SHA for 56d19a5

      View code at this pointBrowse repository at this point
18. Commits on Aug 4, 2025

    • Revert "Migrate ScalarType to headeronly (#159416)"

      Show description for 7e8197e
      pytorchmergebot

      committed
      7e8197eView commit details

      Copy full SHA for 7e8197e

      View code at this pointBrowse repository at this point
19. Commits on Aug 3, 2025

    • [BE][MPS] Build metal kernels of MacOS-14+ (#159733)

      Show description for e2a5c42
      

      malfet

      authored and

      pytorchmergebot

      committed
      e2a5c42View commit details

      Copy full SHA for e2a5c42

      View code at this pointBrowse repository at this point
20. Commits on Aug 1, 2025

    • Remove unused input parameter in ExpandableSegment (#159356)

      Show description for c0e0126
      

      guangyey

      authored and

      pytorchmergebot

      committed
      c0e0126View commit details

      Copy full SHA for c0e0126

      View code at this pointBrowse repository at this point

    • Migrate ScalarType to headeronly (#159416)

      Show description for 1371a98
      

      janeyx99

      authored and

      pytorchmergebot

      committed
      1371a98View commit details

      Copy full SHA for 1371a98

      View code at this pointBrowse repository at this point

    • Revert "Refactor CUDAAllocatorConfig to reuse AcceleratorAllocatorConfig (#150312)"

      Show description for 5cc6a0a
      pytorchmergebot

      committed
      5cc6a0aView commit details

      Copy full SHA for 5cc6a0a

      View code at this pointBrowse repository at this point