Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Jan 23, 2025

• [ONNX] Remove LegacyDynamoStrategy (#145442)

  Show description for f52901a
  

  justinchuby

  authored and

  pytorchmergebot

  committed
  f52901a

  Copy full SHA for f52901a

• [BE] Remove test_modules from FIXME_inductor_dont_reset_dynamo (#145306)

  Show description for 28c251d
  

  masnesral

  authored and

  pytorchmergebot

  committed
  28c251d

  Copy full SHA for 28c251d

• [c10/metal] Add a vectype variant for `short`/`int`/`long` (#145430)

  Show description for f56c638
  

  dcci

  authored and

  pytorchmergebot

  committed
  f56c638

  Copy full SHA for f56c638

• [dynamo][dicts] Insert LENTGH guard on an if condition on dict (#145432)

  Show description for c581981
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  c581981

  Copy full SHA for c581981

• [ROCm] CK SDPA - Move arch check to CK patch (#144777)

  Show description for faa10fa
  

  alugorey

  authored and

  pytorchmergebot

  committed
  faa10fa

  Copy full SHA for faa10fa

• [c10] catch c10 error and log message (#145413)

  Show description for 5e6451e
  

  c-p-i-o

  authored and

  pytorchmergebot

  committed
  5e6451e

  Copy full SHA for 5e6451e

• Generalize pin memory logic for accelerator when non blocking copy happened (#143783)

  Show description for 719938c
  

  guangyey

  authored and

  pytorchmergebot

  committed
  719938c

  Copy full SHA for 719938c

• Introduce a new API isAcceleratorExcluded (#144959)

  Show description for 28b6430
  

  guangyey

  authored and

  pytorchmergebot

  committed
  28b6430

  Copy full SHA for 28b6430

• [dynamo] Support fx map_aggregate (#145351)

  Show description for 5a18f1e
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  5a18f1e

  Copy full SHA for 5a18f1e

• Revert "Align CPU behavior with CUDA for `ConvTranspose` when `out_channels=0` (#142859)"

  Show description for d95a6ba
  pytorchmergebot

  committed
  d95a6ba

  Copy full SHA for d95a6ba

• Move privateuse1 test out of test_utils and make them serial (#145380)

  Show description for 0d28188
  

  albanD

  authored and

  pytorchmergebot

  committed
  0d28188

  Copy full SHA for 0d28188

• [ROCm] Update rocm.yml and add rocm-mi300.yml (#145398)

  Show description for c9e12d6
  

  2 peopleauthored and

  pytorchmergebot

  committed
  c9e12d6

  Copy full SHA for c9e12d6

• Improve softmax's perf in cuda (#144679)

  Show description for 1e32842
  

  wenqiny

  authored and

  pytorchmergebot

  committed
  1e32842

  Copy full SHA for 1e32842

Commits on Jan 22, 2025

• [BE][export] Change custom_op registeration style (#145315)

  Show description for d0a2e11
  

  yiming0416

  authored and

  pytorchmergebot

  committed
  d0a2e11

  Copy full SHA for d0a2e11

• [S481486] Move MTIA dynamic library loading from __init__.py to a separate module (#145322)

  Show description for 4803e20
  

  chaos5958

  authored and

  pytorchmergebot

  committed
  4803e20

  Copy full SHA for 4803e20

• Fix for failure in D68425364 (#145304)

  Show description for 35c8c31
  

  aorenste

  authored and

  pytorchmergebot

  committed
  35c8c31

  Copy full SHA for 35c8c31

• [PyTorch] Add backend aot_eager_decomp_partition_with_mode (#143250)

  Show description for e6a84be
  

  silverlakeli

  authored and

  pytorchmergebot

  committed
  e6a84be

  Copy full SHA for e6a84be

• Revert "Output of nonzero is transposed, fix fake tensor (#144695)"

  Show description for f0a210b
  pytorchmergebot

  committed
  f0a210b

  Copy full SHA for f0a210b

• [CUDA][cuBLAS] Add fp16 accumulate option to cuBLAS/cuBLASLt (#144441)

  Show description for de945d7
  

  eqy

  authored and

  pytorchmergebot

  committed
  de945d7

  Copy full SHA for de945d7

• Revert "[BE]: Simplify set add with set update (#145152)"

  Show description for 6e53588
  pytorchmergebot

  committed
  6e53588

  Copy full SHA for 6e53588

• Revert "Enable grep_linter to use -a (#144589)"

  Show description for dddf52b
  pytorchmergebot

  committed
  dddf52b

  Copy full SHA for dddf52b

• [compiled autograd] support Tensor Subclasses in AOTBackward (#144115)

  Show description for 082c28c
  

  zou3519

  authored and

  pytorchmergebot

  committed
  082c28c

  Copy full SHA for 082c28c

• [compiled autograd] stop specializing on metadata during initial trace (#143417)

  Show description for 99dd1bf
  

  zou3519

  authored and

  pytorchmergebot

  committed
  99dd1bf

  Copy full SHA for 99dd1bf

• [compiled autograd] Always proxy autograd.Function nodes; handle AOT backwards (#143405)

  Show description for ec820fe
  

  zou3519

  authored and

  pytorchmergebot

  committed
  ec820fe

  Copy full SHA for ec820fe

• [compiled autograd] Proxy nodes for user-defined C++ torch::autograd::Function (#143387)

  Show description for 784bb21
  

  zou3519

  authored and

  pytorchmergebot

  committed
  784bb21

  Copy full SHA for 784bb21

• [compiled autograd] Proxy a node for CopyBackwards into the graph (#143304)

  Show description for 8c7c5f7
  

  zou3519

  authored and

  pytorchmergebot

  committed
  8c7c5f7

  Copy full SHA for 8c7c5f7

• [compiled autograd] Proxy opaque nodes for built-in autograd nodes (#143296)

  Show description for 5531faf
  

  zou3519

  authored and

  pytorchmergebot

  committed
  5531faf

  Copy full SHA for 5531faf

• [BE]: Simplify set add with set update (#145152)

  Show description for 0cb9b22
  

  2 peopleauthored and

  pytorchmergebot

  committed
  0cb9b22

  Copy full SHA for 0cb9b22

• [dynamo] Fix numpy test accuracy error induced by randomness divergence (#145293)

  Show description for 9f15078
  

  StrongerXi

  authored and

  pytorchmergebot

  committed
  9f15078

  Copy full SHA for 9f15078

• Binary upload checksum (#144887)

  Show description for 2efa98d
  

  clee2000

  authored and

  pytorchmergebot

  committed
  2efa98d

  Copy full SHA for 2efa98d

• [NVIDIA] Jetson Thor Blackwell Support codegen (#145395)

  Show description for a57133e
  

  johnnynunez

  authored and

  pytorchmergebot

  committed
  a57133e

  Copy full SHA for a57133e

• Reverting the PR adding Kleidiai-based int4 kernels (#145392)

  Show description for 0940eb6
  

  albanD

  authored and

  pytorchmergebot

  committed
  0940eb6

  Copy full SHA for 0940eb6

• [Doc] Add period at the end of the sentence (#145384)

  Show description for 95ff9f0
  

  malfet

  authored and

  pytorchmergebot

  committed
  95ff9f0

  Copy full SHA for 95ff9f0

• [audio hash update] update the pinned audio hash (#145328)

  Show description for 3917053
  

  pytorchupdatebot

  authored and

  pytorchmergebot

  committed
  3917053

  Copy full SHA for 3917053

• [MPSInductor] Add `gamma` op (#145341)

  Show description for 70ccbad
  

  malfet

  authored and

  pytorchmergebot

  committed
  70ccbad

  Copy full SHA for 70ccbad

Pagination

Previous

Next