platform/upstream/pytorch.git
2021-08-12 Zsolt Dollenstein[codemod][lint][fbcode/c*] Enable BLACK by default
2021-08-12 Kushashwa Ravi... [reland] OpInfo: `adaptive_avg_pool2d` (#62935)
2021-08-12 Rong Rong ... [BE] shorten CI name part2 (#63030)
2021-08-12 Rohan VarmaSkip zero test on windows (#63087)
2021-08-12 Peter BellBatchNorm: Use resize_output and empty, instead of...
2021-08-12 Supriya Rao[quant] Make version 1 the default for get_default_qat_...
2021-08-12 Pritam DamaniaFix sharded tensor tests. (#63054)
2021-08-12 Meghan LelePort `log_softmax_backward_data` to structured kernel...
2021-08-12 Meghan LelePort `log_softmax` to structured kernel (#57374)
2021-08-12 zhouzhuojieAdd ciflow_ruleset.json generator along with gha ci...
2021-08-11 Jiewen TanImprove IMethod::getArgumentNames to deal with empty...
2021-08-11 Amy HeFix Nnapi backend execute's dangling pointer (#63092)
2021-08-11 Nikita ShulgaFix warnings (#62930)
2021-08-11 Tao Xu[iOS][GPU] Consolidate array and non-array kernel for...
2021-08-11 Richard Barnesirange-ify 13b (#62476)
2021-08-11 CaoEAdd BFloat16 support for unique and unique_consecutive...
2021-08-11 Alexander GrundAdd Github action to upload full source releases (...
2021-08-11 Xiang GaoEmbedding thrust->cub: unique (#63042)
2021-08-11 Howard Cheng[PyTorch] Add flop count for addmm (#61895)
2021-08-11 Salil DesaiXNNPack Input Pointer Caching Comment (#62818)
2021-08-11 rusty1s`_convert_coo_to_csr` CPP and CUDA functionality (...
2021-08-11 Pritam DamaniaAdd a _RemoteDevice structure for ShardedTensor/Shardin...
2021-08-11 Jacob Szwejbka[Pytorch Edge] Move RuntimeCompatibilityInfo Factory...
2021-08-11 Stephen Macke[easy] add an `inplace` argument to MutableNetProto...
2021-08-11 zhouzhuojieFix gha render-test-result mixed failure passthrough...
2021-08-11 Yida WangFix issues with printing certain torch modules (#62447)
2021-08-11 Peter BellShard python_functions.cpp (#62186)
2021-08-11 Sze Wai Celeste... Fix inconsisteny between Python and JIT power operation...
2021-08-11 Dmytro DzhulgakovFix CUDA_KERNEL_ASSERT ambiguous symbol in NDEBUG mode...
2021-08-11 Pritam Damania[4/N] Enable opt-asan for distributed unit tests. ...
2021-08-11 Lu FangBack out "[fx] store Tracer class on Graph and GraphMod...
2021-08-11 Rishi Purirebase for autocast updates to include device_type...
2021-08-11 Wei-Sheng ChinFix missing element types and shapes when autograd...
2021-08-11 Natalia Gimelsheinremove dead code (#63031)
2021-08-11 Natalia GimelsheinRevert D30199482: [pytorch][PR] Add BFloat16 support...
2021-08-11 Richard BarnesUse `const auto` with irange (#62990)
2021-08-11 Eddie Yanchange nccl version reporting (#62916)
2021-08-10 tktrungnaUpdate test_torch_deploy (#62838)
2021-08-10 tktrungnaupdate test_libtorch (#62797)
2021-08-10 tktrungnaupdate test distributed (#62796)
2021-08-10 tktrungnaupdate test_vulkan (#62795)
2021-08-10 tktrungnaupdate test_rpc (#62781)
2021-08-10 Matej Sladek[ONNX] add support for prim::Unitialized in lower_tuple...
2021-08-10 Howard HuangRemove process_group_agent and faulty_process_group_age...
2021-08-10 Natalia Gimelsheinfix sort and topk with discontiguous out (#63029)
2021-08-10 Hanton Yang[iOS] enable Metal in the nightly build (#62855)
2021-08-10 Christian Puhrschtest_cudnn_convolution_relu skipCUDAIfRocm
2021-08-10 Victor QuachAdd docstring for saved tensors default hooks (#62361)
2021-08-10 Tao Xu[iOS][CI] Store every version of nightlies in S3 (...
2021-08-10 Jerry Zhang[quant][graphmode] Reference pattern support for elu...
2021-08-10 kshitij12345[fix] TestMultiThreadAutograd: propagate exception...
2021-08-10 Amy He[1/N] Nnapi backend execute and compile (#62272)
2021-08-10 CaoEAdd BFloat16 support for unique and unique_consecutive...
2021-08-10 Jerry Zhang[quant][refactor] Checking activation_dtype instead...
2021-08-10 Raghav KansalLU solve uses cuBLAS and cuSOLVER for matrices with...
2021-08-10 Wanchao Liang[sharded_tensor] add default fields to ShardedTensorMet...
2021-08-10 Rohan Varma[DDP] Dont set thread local state in reducer autograd...
2021-08-10 Pyre Bot Jr[typing] suppress errors in `fbcode/caffe2` - batch 2
2021-08-10 Elias EllisonTest shape analysis with opinfos (#59814)
2021-08-10 Elias Ellisonadd ssupport for a few more opinfos in jit (#59812)
2021-08-10 Elias EllisonDon't substitute in symbolic shapes to shape compute...
2021-08-10 Elias Ellisonsmall cleanups (#59810)
2021-08-10 Elias EllisonOnly optimize after change (redo) (#59809)
2021-08-10 Michael Suo[jit] warn if _check_overload_body fails to find source
2021-08-10 Supriya Rao[quant] Update get_default_qat_qconfig to return the...
2021-08-10 Supriya Rao[quant] add reduce_range option to FusedMovingAvgFakeQu...
2021-08-10 Peter BellCodegen: Fix operator::name on windows (#62278)
2021-08-10 Edward YangReject kwonly arguments passed positionally in torch...
2021-08-10 Sameer DeshmukhAllow LocalResponseNorm to accept 0 dim batch sizes...
2021-08-10 Luca WehrstedtUpdate TensorPipe submodule
2021-08-10 Rohan Varma[Reland][DDP] Support not all outputs used in loss...
2021-08-10 Ilqar RamazanliTo fix variance computation for complex Adam (#62946)
2021-08-09 Jerry Zhang[quant][graphmode][fx] Attach a weight qparam dict...
2021-08-09 zhouzhuojieSimplify the logic of running ci workflow codegen ...
2021-08-09 Richard Barnesirange-ify 12b (#62484)
2021-08-09 Peter BellShard Operators.cpp (#62185)
2021-08-09 Richard Barnesirange-ify 13d (#62477)
2021-08-09 peterjc123Enable rebuilds for Ninja on Windows (#62948)
2021-08-09 Marjan FariborzBFP16 quantization/dequantization (#62974)
2021-08-09 Xiang GaoMigrate Embedding thrust sort to cub sort (#62495)
2021-08-09 =Use output memory format based on input for cudnn_convo...
2021-08-09 Richard Barnesirange-ify 12 (#62120)
2021-08-09 Richard Barnesirange-ify 1 (#62193)
2021-08-09 zhouzhuojieFix render_test_results if condition on always() (...
2021-08-09 Rohan Varma[reland] Gate DistributedOptimizers on RPC availability...
2021-08-09 Richard Barnesirange-ify 8d (#62505)
2021-08-09 Bradley Davis[fx] store Tracer class on Graph and GraphModule for...
2021-08-09 Nikita ShulgaMark unused functions with `C10_UNUSED` (#62929)
2021-08-09 peterjc123Stop exporting symbols in anonymous namespaces (#62952)
2021-08-09 Mike Iovine[Static Runtime] Add tests for all aten ops (#62347)
2021-08-09 Zeina Migeedhandle get_attr opearations in typechecker (#62682)
2021-08-09 Bert MaherLinker version script to hide LLVM symbols (#62906)
2021-08-09 Andrew GuAdd ``allow_empty_param_list`` to functional optimizers...
2021-08-09 Sangbaek Park[Vulkan] Added Hardshrink op (#62870)
2021-08-09 Zeina MigeedChange output node handling for typechecker to deal...
2021-08-09 Edward Yang__torch_dispatch__: Populate kwargs dictionary with...
2021-08-09 Jane XuModify GHA CI to use PYTORCH_IGNORE_DISABLED_ISSUES...
2021-08-09 Raghavan Raman[Static Runtime] Added a cache for NNC generated code...
2021-08-09 Rong Rong ... Enable upper for torch.linalg.cholesky (#62434)
2021-08-09 Raghavan Raman[nnc] Updated IR cloning to create clones of expression...
next