summaryrefslogtreecommitdiff
path: root/aten
AgeCommit message (Expand)AuthorFilesLines
2021-08-18[ONNX] Add support for opset14 in PT-ONNX exporter (#59486) (#62758)BowenBao1-1/+2
2021-08-18[pytorch] Make qconv forward() thread safe (#63432)John Shen3-5/+7
2021-08-18Use `fastAtomicAdd` in EmbeddingBag (mode "max") backward (#63298)Masaki Kozuki1-4/+6
2021-08-18Reverting launch bounds change in topK that induced a regression in perf (#63...Rishi Puri1-2/+2
2021-08-18A tiny fix in MT19937RNGEngine (#63219)cyy1-1/+0
2021-08-18add `OpInfo` for `torch.linalg.tensorinv` (#62326)Philip Meier1-6/+3
2021-08-17`F.avg_pool3` CUDA backward: gpuAtomicAddNoReturn -> fastAtomicAdd (#63387)Masaki Kozuki1-3/+5
2021-08-17Fix zero-dim handling in torch.matmul (#63359)Richard Zou1-1/+6
2021-08-17[vulkan] Add log_softmax (#63193)Ivan Kobzarev3-3/+88
2021-08-16Cleanup dead code (#63328)Peter Bell1-55/+0
2021-08-16Workaround for cuFFT bug (#63327)Peter Bell1-1/+33
2021-08-16Refactor NnapiCompilation registration into it's own file (#63183)Amy He2-27/+21
2021-08-16Make `torch.lu` differentiable for wide/tall inputs + jit (#61564)Nikita Vedeneev1-1/+1
2021-08-13[Pytorch Edge] Enable kineto profiler on mobile via EdgeKinetoProfiler (#62419)Kimish Patel1-5/+7
2021-08-13[Pytorch Profiler] Add debug_handles to KinetoEvent (#62228)Kimish Patel1-0/+36
2021-08-13add substract of max and testcase (#63132)leslie-fang-intel1-3/+2
2021-08-13[ROCm] Update HIP_VERSION to TORCH_HIP_VERSION (#62786)Pruthvi Madugundu6-16/+12
2021-08-13Allow Average Pooling modules to accept tensors with 0-dim batch sizes. (#62025)Sameer Deshmukh4-22/+48
2021-08-13ENH Adds no_batch_dim to FractionalMaxPool2d (#62490)Thomas J. Fan1-1/+1
2021-08-13Port `norm` kernel to structured kernels. (#62711)Yukio Siraichi2-97/+107
2021-08-13Port `prod` kernel to structured kernels. (#62024)Yukio Siraichi2-72/+86
2021-08-13Port `mean` kernel to structured kernels. (#61643)Yukio Siraichi3-29/+47
2021-08-13Make saved tensors default hooks thread local (#62909)Victor Quach4-0/+64
2021-08-13Allow 0-dim batch sizes for AdaptiveMaxPool and MaxPool. (#62088)Sameer Deshmukh10-38/+78
2021-08-12ENH Migrate nll_loss2d from THC to ATen (#62826)Thomas J. Fan8-653/+527
2021-08-12[iOS][GPU] Fix the clamp shader function for x86_64 (#63062)Tao Xu5-46/+42
2021-08-12LayerNorm Support in autodiff: (#50467)jiej2-2/+2
2021-08-11BatchNorm: Use resize_output and empty, instead of empty_like (#63084)Peter Bell3-3/+15
2021-08-11[quant] Make version 1 the default for get_default_qat_qconfig (#63043)Supriya Rao2-32/+75
2021-08-11Port `log_softmax_backward_data` to structured kernel (#62372)Meghan Lele7-99/+174
2021-08-11Port `log_softmax` to structured kernel (#57374)Meghan Lele7-62/+125
2021-08-11Fix warnings (#62930)Nikita Shulga3-13/+18
2021-08-11[iOS][GPU] Consolidate array and non-array kernel for upsampling_nearest2d (#...Tao Xu5-31/+43
2021-08-11Add BFloat16 support for unique and unique_consecutive on CPU (#62559)CaoE1-7/+15
2021-08-11Embedding thrust->cub: unique (#63042)Xiang Gao6-46/+28
2021-08-11XNNPack Input Pointer Caching Comment (#62818)Salil Desai1-0/+11
2021-08-11`_convert_coo_to_csr` CPP and CUDA functionality (#61838)rusty1s3-0/+121
2021-08-10rebase for autocast updates to include device_type and dtype flags (#61002)Rishi Puri2-1/+15
2021-08-10remove dead code (#63031)Natalia Gimelshein5-233/+0
2021-08-10Revert D30199482: [pytorch][PR] Add BFloat16 support for unique and unique_co...Natalia Gimelshein1-15/+7
2021-08-10fix sort and topk with discontiguous out (#63029)Natalia Gimelshein2-4/+18
2021-08-10[1/N] Nnapi backend execute and compile (#62272)Amy He2-169/+199
2021-08-10Add BFloat16 support for unique and unique_consecutive on CPU (#62559)CaoE1-7/+15
2021-08-10LU solve uses cuBLAS and cuSOLVER for matrices with dim > 1024 (#61815)Raghav Kansal1-2/+5
2021-08-09Shard Operators.cpp (#62185)Peter Bell1-0/+2
2021-08-09Migrate Embedding thrust sort to cub sort (#62495)Xiang Gao3-65/+80
2021-08-09Use output memory format based on input for cudnn_convolution_relu (#62482)=1-1/+1
2021-08-09irange-ify 1 (#62193)Richard Barnes1-2/+0
2021-08-09Mark unused functions with `C10_UNUSED` (#62929)Nikita Shulga10-25/+25
2021-08-09[Vulkan] Added Hardshrink op (#62870)Sangbaek Park4-1/+212
generated by cgit v1.2.3 (git 2.25.1) at 2024-12-01 09:06:54 +0000