prs_0.4

https://github.com/pytorch/pytorch/pull/6692 Update from Facebook
https://github.com/pytorch/pytorch/pull/6681 fix broken code from rebasing
https://github.com/pytorch/pytorch/pull/6677 Make torch.backends.mkl.is_available() work without importing
https://github.com/pytorch/pytorch/pull/6673 Scope variables inside the dataloader
https://github.com/pytorch/pytorch/pull/6670 Update tensors.rst Tensor introduction
https://github.com/pytorch/pytorch/pull/6665 Fix LSTM and GRU parameters description
https://github.com/pytorch/pytorch/pull/6664 Adding dispatch to Tensors
https://github.com/pytorch/pytorch/pull/6659 Fix some loss output sizes
https://github.com/pytorch/pytorch/pull/6656 randperm supports n=0
https://github.com/pytorch/pytorch/pull/6654 Always compute gradients for the gradcheck inputs
https://github.com/pytorch/pytorch/pull/6648 Support gpu triangle solve
https://github.com/pytorch/pytorch/pull/6647 Gltensor fix
https://github.com/pytorch/pytorch/pull/6642 Allow traces to call @script functions
https://github.com/pytorch/pytorch/pull/6641 Codemod to update our codebase to 0.4 standard
https://github.com/pytorch/pytorch/pull/6637 Bind 0-dim variables without requires grad to int64/double similar to
https://github.com/pytorch/pytorch/pull/6629 Add Module.to
https://github.com/pytorch/pytorch/pull/6628 Make dtype in .to positional rather than kwarg only
https://github.com/pytorch/pytorch/pull/6617 nll_loss: Fixed text of error message in case of unexpected target size
https://github.com/pytorch/pytorch/pull/6616 [Adagrad optimization] adding initial_accumulator_value parameter to Adagrad
https://github.com/pytorch/pytorch/pull/6599 Split set_default_tensor_type(dtype) into set_default_dtype(dtype).
https://github.com/pytorch/pytorch/pull/6592 Change to ldd parsing regex
https://github.com/pytorch/pytorch/pull/6588 Add tensor.to(device) method.
https://github.com/pytorch/pytorch/pull/6575 Fix torch.nn.RNN parameters description
https://github.com/pytorch/pytorch/pull/6573 Add dtypes (with reasonable defaults) to sum, prod, cumsum, cumprod.
https://github.com/pytorch/pytorch/pull/6553 Restore allow_unused functionality
https://github.com/pytorch/pytorch/pull/6541 Fix regression that STFT has no backward.
https://github.com/pytorch/pytorch/pull/6534 Conda binary changes
https://github.com/pytorch/pytorch/pull/6528 Support arbitrary number of batch dimensions in *FFT
https://github.com/pytorch/pytorch/pull/6517  More precise digamma
https://github.com/pytorch/pytorch/pull/6507 Adding autofunction entry for torch.randint
https://github.com/pytorch/pytorch/pull/6485 Add SmallVector from llvm
https://github.com/pytorch/pytorch/pull/6484 Sync current changes in ACL backend
https://github.com/pytorch/pytorch/pull/6470 Separate cuda-ness from dtype.
https://github.com/pytorch/pytorch/pull/6467 [Re-checkpointing] Autograd container for trading compute for memory
https://github.com/pytorch/pytorch/pull/6463 [pytorch] Fix signed random_
https://github.com/pytorch/pytorch/pull/6438 Fix reflection padding boundary checks
https://github.com/pytorch/pytorch/pull/6426 Slice (instead of copy) when indexing by a zero-dim tensor
https://github.com/pytorch/pytorch/pull/6425 bottleneck supports better user-provided arguments
https://github.com/pytorch/pytorch/pull/6420 Use string comparison in OS check
https://github.com/pytorch/pytorch/pull/6418 [pytorch] Fix clamp is missing kwarg out (#6028)
https://github.com/pytorch/pytorch/pull/6409 Fix incorrect error message in convolution_expand_param_if_needed
https://github.com/pytorch/pytorch/pull/6405 Quote arguments only when possible
https://github.com/pytorch/pytorch/pull/6401 Add CUDA headers
https://github.com/pytorch/pytorch/pull/6396 Fixes #6386, Use copies instead of symbolic files
https://github.com/pytorch/pytorch/pull/6371 refactor reduce arg to _Loss superclass
https://github.com/pytorch/pytorch/pull/6367 Fix activation images not showing up on official website
https://github.com/pytorch/pytorch/pull/6358 Fix Sphinx's incorrect rendering of arg type torch.dtype
https://github.com/pytorch/pytorch/pull/6341 Allow script_methods to be defined out of order
https://github.com/pytorch/pytorch/pull/6330 Add missing arguments to the GivenTensorFill schema
https://github.com/pytorch/pytorch/pull/6327 Add total_length option to pad_packed_sequence
https://github.com/pytorch/pytorch/pull/6314 Move instruction set specific code to anonymous namespace
https://github.com/pytorch/pytorch/pull/6307 Implement torch.einsum (fixes #1889)
https://github.com/pytorch/pytorch/pull/6293 Remove eigen impl for arg_max and arg_min
https://github.com/pytorch/pytorch/pull/6289 Add default args to loss functions in native_functions.yaml
https://github.com/pytorch/pytorch/pull/6283 Add string-style devices to all tensors.
https://github.com/pytorch/pytorch/pull/6281 Use reshape({-1})
https://github.com/pytorch/pytorch/pull/6274 Add a CODEOWNERS file
https://github.com/pytorch/pytorch/pull/6272 [ready] Implement log2 and log10 in PyTorch
https://github.com/pytorch/pytorch/pull/6254 Change Python Arg Parser to only read default params if they are assigned
https://github.com/pytorch/pytorch/pull/6252 fix assertion error when input size smaller than number of module_copies
https://github.com/pytorch/pytorch/pull/6251 Fix Tensor.__setstate__ for legacy Tensor state
https://github.com/pytorch/pytorch/pull/6250 Remove unnecessary properties from Layout.
https://github.com/pytorch/pytorch/pull/6249 Add arg checks in torch.utils.data.Sampler classes
https://github.com/pytorch/pytorch/pull/6244 Fix SGD lr check failing on default value
https://github.com/pytorch/pytorch/pull/6242 Fix potential UB when input is empty
https://github.com/pytorch/pytorch/pull/6238 Print the diff files to aid in debugging when it's wrong.
https://github.com/pytorch/pytorch/pull/6236 Remove unused variable in Layout.cpp.
https://github.com/pytorch/pytorch/pull/6232 Detect re-initialization of _C shared library
https://github.com/pytorch/pytorch/pull/6230 Fix memory leak in maxpool3d backwards
https://github.com/pytorch/pytorch/pull/6229 Fix sharing of empty tensor in multiprocessing
https://github.com/pytorch/pytorch/pull/6221 Fix AvgPool breaking changes
https://github.com/pytorch/pytorch/pull/6211 Fix sparse embedding backward when input contains only padding_idx
https://github.com/pytorch/pytorch/pull/6207 Fix argument checking for inlining a module
https://github.com/pytorch/pytorch/pull/6199 use peephole 'optimization' to export default hidden size with correct semantics
https://github.com/pytorch/pytorch/pull/6192 avx_mathfun.h is imprecise
https://github.com/pytorch/pytorch/pull/6185 Tell source users about TORCH_CUDA_ARCH_LIST
https://github.com/pytorch/pytorch/pull/6173 Update torch.nn.init and torch.nn.utils.clip_grad
https://github.com/pytorch/pytorch/pull/6159 Move helper scripts to new repo
https://github.com/pytorch/pytorch/pull/6158 Add dtype arg to torch.*_window; Add dtype.is_floating_point
https://github.com/pytorch/pytorch/pull/6151 Delete NNPACK
https://github.com/pytorch/pytorch/pull/6146 Fix logic inside insertInput
https://github.com/pytorch/pytorch/pull/6145 Introduce torch.layout and split layout from dtypes.
https://github.com/pytorch/pytorch/pull/6136 [WIP] randint function
https://github.com/pytorch/pytorch/pull/6128 Make precision matrix computation in mvn stable
https://github.com/pytorch/pytorch/pull/6118 Fix fft when any of the input dimensions is not aligned
https://github.com/pytorch/pytorch/pull/6114 Avoid generating torch.*_backward_(input|weight|bias)
https://github.com/pytorch/pytorch/pull/6113 Support returning dictionaries in DataParallel
https://github.com/pytorch/pytorch/pull/6110 Fix bilinear performance regression
https://github.com/pytorch/pytorch/pull/6108 Set dataloader.batch_size = None when batch_sampler is given
https://github.com/pytorch/pytorch/pull/6093 Add underscore to nn.init.* and deprecate the original ones
https://github.com/pytorch/pytorch/pull/6089 Update FFT comments
https://github.com/pytorch/pytorch/pull/6088 Unify handling of type_dispatched_args in gen_python_functions.
https://github.com/pytorch/pytorch/pull/6086 Add class-specific error when key mismatch in load_state_dict
https://github.com/pytorch/pytorch/pull/6081 Remove dtypes from legacy tensor.new(...)
https://github.com/pytorch/pytorch/pull/6078 Exp, log, sin, cos vectorized
https://github.com/pytorch/pytorch/pull/6076 Correct argument misspelling.
https://github.com/pytorch/pytorch/pull/6072 NLLLoss: error message for mismatched input/target batch sizes
https://github.com/pytorch/pytorch/pull/6070 Relax constraints on return statements in the script
https://github.com/pytorch/pytorch/pull/6069 Fix printing of unknown binop operator in torchscript
https://github.com/pytorch/pytorch/pull/6062 Enable MKLDNN convolution forward and backward
https://github.com/pytorch/pytorch/pull/6059 Add source location information to error messages
https://github.com/pytorch/pytorch/pull/6058 Create safe and unsafe versions of sparse_coo_tensor
https://github.com/pytorch/pytorch/pull/6043 Update cpuinfo to d0222b4
https://github.com/pytorch/pytorch/pull/6038 Enable TensorDataset to get any number of tensors
https://github.com/pytorch/pytorch/pull/6037 Fix use-after-free bug in peephole pass
https://github.com/pytorch/pytorch/pull/6033 Add additional script module functionality
https://github.com/pytorch/pytorch/pull/6031 Block set from param_group['params']
https://github.com/pytorch/pytorch/pull/6026 Speed up sum over a dimension
https://github.com/pytorch/pytorch/pull/6025 Reorganize third-party libraries into top-level third_party directory
https://github.com/pytorch/pytorch/pull/6023 Fix instance norm
https://github.com/pytorch/pytorch/pull/6022 mkl-include is not installable if your conda is too old.
https://github.com/pytorch/pytorch/pull/6017 Update FAQ to make more sense after tensor/variable merge
https://github.com/pytorch/pytorch/pull/6005 Extra comment about backward vs. grad in engine.
https://github.com/pytorch/pytorch/pull/6004 Move .jenkins to .jenkins/pytorch
https://github.com/pytorch/pytorch/pull/6000 Added parameter range checks for all optimizers
https://github.com/pytorch/pytorch/pull/5997 Add numpy.array-like type inference to torch.tensor.
https://github.com/pytorch/pytorch/pull/5992 fix bias size assert
https://github.com/pytorch/pytorch/pull/5991 add mkl dependencies to setup
https://github.com/pytorch/pytorch/pull/5984 Add pip mkl-devel to the error message about mkl header files
https://github.com/pytorch/pytorch/pull/5980 Support batch LowerCholeskyTransform
https://github.com/pytorch/pytorch/pull/5971 Fix crash when cat-ing empty cuda tensors
https://github.com/pytorch/pytorch/pull/5968 Group Normalization
https://github.com/pytorch/pytorch/pull/5965 Remove pragma once from cpp file
https://github.com/pytorch/pytorch/pull/5955 Recommend citation
https://github.com/pytorch/pytorch/pull/5951 Store perf numbers in S3
https://github.com/pytorch/pytorch/pull/5945 Fix tensor.permute(dims) backward for negative dims
https://github.com/pytorch/pytorch/pull/5936 Add support for printing extra information in Module and refactor redundant codes
https://github.com/pytorch/pytorch/pull/5934 Fix index out of range error when view a scalar as 1-dim tensor
https://github.com/pytorch/pytorch/pull/5928 Remove consumed_input
https://github.com/pytorch/pytorch/pull/5927 Linearly interpolating upsampling fix
https://github.com/pytorch/pytorch/pull/5926 parallel_for_2d fix and guarding avx/avx2 compilation
https://github.com/pytorch/pytorch/pull/5919 add mpi support for DDP
https://github.com/pytorch/pytorch/pull/5914 Fix linking issue in libtorch under macOS
https://github.com/pytorch/pytorch/pull/5913 Optimize unique sorting by using std::vector+sort instead of std::set
https://github.com/pytorch/pytorch/pull/5909 Revert "Fix ImportError with requests in model_zoo"
https://github.com/pytorch/pytorch/pull/5906 Fix integer overflow in remainder operator
https://github.com/pytorch/pytorch/pull/5896 Fix ImportError with requests in model_zoo
https://github.com/pytorch/pytorch/pull/5894 Better error msg for missing mkl headers
https://github.com/pytorch/pytorch/pull/5893 Fix softmax symbolic
https://github.com/pytorch/pytorch/pull/5892 Revert "Enable resetting of batchnorm running stats and cumulative ("simple") moving average"
https://github.com/pytorch/pytorch/pull/5890 Add support for subscripts in Python frontend
https://github.com/pytorch/pytorch/pull/5889 Support legacy empty tensor behavior in cat
https://github.com/pytorch/pytorch/pull/5880 Don't modify requires_grad when running DataParallel in no_grad mode
https://github.com/pytorch/pytorch/pull/5866 Delete stubs from one more place.
https://github.com/pytorch/pytorch/pull/5856 [fft][2 of 3] Forward for fft methods
https://github.com/pytorch/pytorch/pull/5850 Fix crash in new cuda tensor with numpy array
https://github.com/pytorch/pytorch/pull/5846 Softmax symbolic should account for negative dim
https://github.com/pytorch/pytorch/pull/5843 Add support for number and list literals in Python frontend
https://github.com/pytorch/pytorch/pull/5840 Fix nvprof parsing
https://github.com/pytorch/pytorch/pull/5838 Add operator[](int64_t) overload
https://github.com/pytorch/pytorch/pull/5829 fix detach in place error in DDP
https://github.com/pytorch/pytorch/pull/5827 Implement range for loop in script
https://github.com/pytorch/pytorch/pull/5824 introduce shape_as_tensor and reshape_from_tensor_shape
https://github.com/pytorch/pytorch/pull/5822 Make static state function-local
https://github.com/pytorch/pytorch/pull/5820 Namespaced symbols
https://github.com/pytorch/pytorch/pull/5819 Fix error message for cat-ing zero-dim tensors
https://github.com/pytorch/pytorch/pull/5818 Revert "introduce size_as_tensor and resize_from_tensor"
https://github.com/pytorch/pytorch/pull/5817 Cleaner solution to the undefined references in RPC
https://github.com/pytorch/pytorch/pull/5815 Fix convolution type mismatch error message
https://github.com/pytorch/pytorch/pull/5814 Fix kldiv backward on CUDA
https://github.com/pytorch/pytorch/pull/5794 Define RPC types out of source
https://github.com/pytorch/pytorch/pull/5792 introduce size_as_tensor and resize_from_tensor
https://github.com/pytorch/pytorch/pull/5786 Add symbolic functions for cumsum and embedding_bag
https://github.com/pytorch/pytorch/pull/5785 Make the tensor type torch.Tensor instead of torch.autograd.Variable
https://github.com/pytorch/pytorch/pull/5782 fused GLU backward
https://github.com/pytorch/pytorch/pull/5781 [REDO] Add torch.sparse_coo_tensor factory.
https://github.com/pytorch/pytorch/pull/5780 Revert "Add torch.sparse_coo_tensor factory."
https://github.com/pytorch/pytorch/pull/5774 improve handling of precision issue in torch.multinomial (solves #4858)
https://github.com/pytorch/pytorch/pull/5773 use std:: math functions
https://github.com/pytorch/pytorch/pull/5766 Enable resetting of batchnorm running stats and cumulative ("simple") moving average
https://github.com/pytorch/pytorch/pull/5764 Support N-D tensors in Bilinear
https://github.com/pytorch/pytorch/pull/5756 Fixes Variable::data() on UndefinedTensor
https://github.com/pytorch/pytorch/pull/5755 Put torch header install back into the install command
https://github.com/pytorch/pytorch/pull/5749 Allow indexing by scalars and zero-dim tensors
https://github.com/pytorch/pytorch/pull/5747 Save self.numel() for backward computation instead of self
https://github.com/pytorch/pytorch/pull/5745 Add torch.sparse_coo_tensor factory.
https://github.com/pytorch/pytorch/pull/5744 Fix bmm memory leak
https://github.com/pytorch/pytorch/pull/5743 Clean up TraceInput
https://github.com/pytorch/pytorch/pull/5726 Attempt to fix #5718.
https://github.com/pytorch/pytorch/pull/5723 tbb set_num_threads
https://github.com/pytorch/pytorch/pull/5722 Add optimization to norm for common norms
https://github.com/pytorch/pytorch/pull/5717 Fix at::optional return type in fusibleExpandTo
https://github.com/pytorch/pytorch/pull/5713 Ensure torch.tensor and Tensor.new_tensor copy numpy data.
https://github.com/pytorch/pytorch/pull/5710 improve occupancy for cuda rngs
https://github.com/pytorch/pytorch/pull/5707 Delete ""_sym literal form.
https://github.com/pytorch/pytorch/pull/5701 Fix error message in nn.functional.convNd and nn.functional.conv_transposeNd
https://github.com/pytorch/pytorch/pull/5682 Fix floor latex rendering
https://github.com/pytorch/pytorch/pull/5680  implement TripletMarginLoss as a native function
https://github.com/pytorch/pytorch/pull/5674 Only allow dense floating-point types as the default tensor type.
https://github.com/pytorch/pytorch/pull/5669 Add device to Tensor.new_tensor.
https://github.com/pytorch/pytorch/pull/5668 Add torch.empty, torch.full and new_ size Tensor factory methods.
https://github.com/pytorch/pytorch/pull/5659 make dimension checker of `scatter_add_` consistent with `scatter_`
https://github.com/pytorch/pytorch/pull/5657 Check value type for register_buffer
https://github.com/pytorch/pytorch/pull/5655 Add gpu guard for broadcast_coalesce
https://github.com/pytorch/pytorch/pull/5646 implement CosineEmbeddingLoss as a native function and add reduce arg
https://github.com/pytorch/pytorch/pull/5645 Non var
https://github.com/pytorch/pytorch/pull/5644 Fix CUDA btrifact error message using wrong info type
https://github.com/pytorch/pytorch/pull/5643 Add additional deprecated overloads with out kwarg
https://github.com/pytorch/pytorch/pull/5642 Unify error checking for tensor.index_copy_
https://github.com/pytorch/pytorch/pull/5640 Revert "implement CosineEmbeddingLoss as a native function and add reduce arg"
https://github.com/pytorch/pytorch/pull/5629 Traceable dispatch for cast methods
https://github.com/pytorch/pytorch/pull/5622 Alias torch.diagonal, torch.diagflat
https://github.com/pytorch/pytorch/pull/5621 Fix compilation with CUDA < 8.0
https://github.com/pytorch/pytorch/pull/5619 Prefix DataLoaderIter with underscore to discourage subclassing
https://github.com/pytorch/pytorch/pull/5614 replace ExportProxy
https://github.com/pytorch/pytorch/pull/5600 Make torch.arange consistent with numpy.arange
https://github.com/pytorch/pytorch/pull/5596 remove legacy workaround for hinge embedding loss reference fn
https://github.com/pytorch/pytorch/pull/5595 allow application of @symbolic decorators without circular imports
https://github.com/pytorch/pytorch/pull/5594 release() does not need to be virtual
https://github.com/pytorch/pytorch/pull/5593 Fix for a confusion around grammar of Maybe
https://github.com/pytorch/pytorch/pull/5583 Allow indexing tensors with both CPU and CUDA tensors
https://github.com/pytorch/pytorch/pull/5582 Use operator.index to convert indices to Python int
https://github.com/pytorch/pytorch/pull/5581 Fix Variable conversion on the way to/from Python
https://github.com/pytorch/pytorch/pull/5576 Support native namespace functions with type dispatch.
https://github.com/pytorch/pytorch/pull/5575 Implement torch.reshape and Tensor.reshape
https://github.com/pytorch/pytorch/pull/5574 Defer shape analysis failures until runtime
https://github.com/pytorch/pytorch/pull/5558 Missed Step
https://github.com/pytorch/pytorch/pull/5555 Add set_grad_enabled as context manager and function
https://github.com/pytorch/pytorch/pull/5540 add: padding_value to `torch.nn.utils.rnn.pad_sequence`
https://github.com/pytorch/pytorch/pull/5530 Add at::optional from https://github.com/akrzemi1/Optional
https://github.com/pytorch/pytorch/pull/5526 Implement pow() for integer types
https://github.com/pytorch/pytorch/pull/5525 Expunge all occurrences of torch._C._VariableFunctions
https://github.com/pytorch/pytorch/pull/5516 Add virtual destructor to SourceLocation
https://github.com/pytorch/pytorch/pull/5512 CharTensor should be signed
https://github.com/pytorch/pytorch/pull/5508 Replace all uses of 'Tensor or Variable' with 'Tensor'
https://github.com/pytorch/pytorch/pull/5505 Some additional clean-ups
https://github.com/pytorch/pytorch/pull/5503 Add per-element unique op for CPU
https://github.com/pytorch/pytorch/pull/5501 Set default amsgrad param in adam optimizer
https://github.com/pytorch/pytorch/pull/5500 Delete unused files
https://github.com/pytorch/pytorch/pull/5498 Fix naming issue in TensorCompare.cpp
https://github.com/pytorch/pytorch/pull/5493 Lower max jobs osx
https://github.com/pytorch/pytorch/pull/5488 Recompute captures after the parameter is updated
https://github.com/pytorch/pytorch/pull/5483 Enable additional tensor types in Gloo backend
https://github.com/pytorch/pytorch/pull/5482 install pytorch into default conda env
https://github.com/pytorch/pytorch/pull/5476 Deprecate variable factory, use torch.tensor instead
https://github.com/pytorch/pytorch/pull/5473 Remove some uses of torch.is_tensor in favor of isinstance
https://github.com/pytorch/pytorch/pull/5471 Reorganize interned strings into categories.
https://github.com/pytorch/pytorch/pull/5467 Check that parsed_args contains enough space for all parameters
https://github.com/pytorch/pytorch/pull/5466 torch.load() / torch.save() support arbitrary file-like object
https://github.com/pytorch/pytorch/pull/5465 Use 'Tensor' instead of 'Variable' in type error messages
https://github.com/pytorch/pytorch/pull/5464 More Variable/Tensor clean-ups
https://github.com/pytorch/pytorch/pull/5453 Added 3d grid sampler (for volumetric transformer networks)
https://github.com/pytorch/pytorch/pull/5447 implement CosineEmbeddingLoss as a native function and add reduce arg
https://github.com/pytorch/pytorch/pull/5444 Add dtype to torch.Tensor constructors and accept them in set_default_tensor_type
https://github.com/pytorch/pytorch/pull/5441 Support type conversion via type(dtype).
https://github.com/pytorch/pytorch/pull/5433 speed up CPU EmbeddingBag (indexSelectAdd op)
https://github.com/pytorch/pytorch/pull/5423 add dependency for OS X
https://github.com/pytorch/pytorch/pull/5419 Introduce torch.tensor (was torch.autograd.variable).
https://github.com/pytorch/pytorch/pull/5418 Avoid extra cpu->cpu copy in dispatch_type.
https://github.com/pytorch/pytorch/pull/5415 Set python random seed in workers
https://github.com/pytorch/pytorch/pull/5414 Empty sparse tensor copy reverses dimI, dimV.
https://github.com/pytorch/pytorch/pull/5413 Remove two uses of the old Tensor class
https://github.com/pytorch/pytorch/pull/5408 Expose gradients w.r.t. input & weight for conv1d, conv2d, conv3d in Python
https://github.com/pytorch/pytorch/pull/5393 [ready] Add logdet and slogdet
https://github.com/pytorch/pytorch/pull/5392 Mark functions that shouldn't end up in torch. as method-only.
https://github.com/pytorch/pytorch/pull/5390 Don't python bind 'tensor' or 'sparse_coo_tensor'.
https://github.com/pytorch/pytorch/pull/5384 Add support for device python arguments with constructors.
https://github.com/pytorch/pytorch/pull/5380 Ignore FileNotFoundError when shutting down in data_queue.get
https://github.com/pytorch/pytorch/pull/5378 Add faq on cuda memory management and dataloder worker seeds
https://github.com/pytorch/pytorch/pull/5371 Fix undefined refence to convolve_5x5_sse on SSE4.1 CPUs
https://github.com/pytorch/pytorch/pull/5366 Fix wrong argument name
https://github.com/pytorch/pytorch/pull/5361 Handle copying empty sparse tensors to/from CPU, GPU.
https://github.com/pytorch/pytorch/pull/5352 Add a disabled-configs.txt interlock.
https://github.com/pytorch/pytorch/pull/5350 Embedding.from_pretrained factory
https://github.com/pytorch/pytorch/pull/5348 Added torch.distributed.launch module for easier multi-proc/node distributed job launching
https://github.com/pytorch/pytorch/pull/5346 Implement MarginRankingLoss as native function and add reduce=True arg to it
https://github.com/pytorch/pytorch/pull/5343 Support dtypes in legacy new constructors.
https://github.com/pytorch/pytorch/pull/5335 Improve sparse variable printing.
https://github.com/pytorch/pytorch/pull/5334 Fix the bug of only processing one attribute
https://github.com/pytorch/pytorch/pull/5328 implement double backwards for MaxPool3d
https://github.com/pytorch/pytorch/pull/5324 Improve CUDA extension support
https://github.com/pytorch/pytorch/pull/5321 Various dtype improvements.
https://github.com/pytorch/pytorch/pull/5320 Make _like dtype arguments keyword only.
https://github.com/pytorch/pytorch/pull/5318 Remove _out variants of like functions.
https://github.com/pytorch/pytorch/pull/5317 add guards when source of container cannot be retreived
https://github.com/pytorch/pytorch/pull/5312 Change output_declarations in function_wrapper.py to be a NamedTuple
https://github.com/pytorch/pytorch/pull/5300 Make ReduceLROnPlateau serializable.
https://github.com/pytorch/pytorch/pull/5299 Raise an error if target is out-of-bounds in ClassNLLCriterion
https://github.com/pytorch/pytorch/pull/5294 Configurable flushing denormal numbers on CPU
https://github.com/pytorch/pytorch/pull/5293 add control flow to interpreter
https://github.com/pytorch/pytorch/pull/5279 Speed-up nn.Linear for the 3d input case
https://github.com/pytorch/pytorch/pull/5277 Use TORCH_EXTENSION_NAME macro to avoid mismatched module/extension name
https://github.com/pytorch/pytorch/pull/5276 Fix  __syncthread in SpatialClassNLLCriterion.cu
https://github.com/pytorch/pytorch/pull/5275 Fixes UB when using legacy python functions and mark_non_differentiable
https://github.com/pytorch/pytorch/pull/5273 Implement torch.isnan
https://github.com/pytorch/pytorch/pull/5255 check attribute existence in torch.legay.nn.SpatialFullConvolution
https://github.com/pytorch/pytorch/pull/5251 Add a FAQ, for now just 'out of memory' advice.
https://github.com/pytorch/pytorch/pull/5250 Revert "Remove unnecessary __syncthreads before reduceBlock"
https://github.com/pytorch/pytorch/pull/5246 Fix call to assertNotEqual
https://github.com/pytorch/pytorch/pull/5245 Add numpy-style dtypes to Variable factories.
https://github.com/pytorch/pytorch/pull/5242 Remove unnecessary __syncthreads before reduceBlock
https://github.com/pytorch/pytorch/pull/5238 CUDA multinomial fix
https://github.com/pytorch/pytorch/pull/5233 Include __delitem__ for Sequential
https://github.com/pytorch/pytorch/pull/5230 Check GCC version on Ubuntu
https://github.com/pytorch/pytorch/pull/5228 Fix for PRId64
https://github.com/pytorch/pytorch/pull/5225 Merge Variable and Tensor classes
https://github.com/pytorch/pytorch/pull/5221 Improve Function interface
https://github.com/pytorch/pytorch/pull/5216 Implement torch.util.bottleneck
https://github.com/pytorch/pytorch/pull/5215 Fix GraphExecutor and add more AD formulas
https://github.com/pytorch/pytorch/pull/5204 Implement symbolic for slice operation
https://github.com/pytorch/pytorch/pull/5203 Additional sparse Variable fixes
https://github.com/pytorch/pytorch/pull/5202 Modest refactor of .jenkins scripts
https://github.com/pytorch/pytorch/pull/5196 Add missing async deprecated wrapper to tools/autograd/templates/pyth…
https://github.com/pytorch/pytorch/pull/5194 CUDA 9
https://github.com/pytorch/pytorch/pull/5184 Make Python autograd functions respect grad mode
https://github.com/pytorch/pytorch/pull/5165 make explicit about keyword-onlyness of `out`
https://github.com/pytorch/pytorch/pull/5162 Replace NULL with nullptr in autograd
https://github.com/pytorch/pytorch/pull/5160 Dropout
https://github.com/pytorch/pytorch/pull/5158 Enable scalars.
https://github.com/pytorch/pytorch/pull/5150 add reduce=True arg to MultiMarginLoss
https://github.com/pytorch/pytorch/pull/5144 Add a new_tensor instance method to Variable that takes only data.
https://github.com/pytorch/pytorch/pull/5143 Adjust stft result comparison precision to 7e-6
https://github.com/pytorch/pytorch/pull/5142 Allow zero-dim tensors to be bound to at::Scalar
https://github.com/pytorch/pytorch/pull/5130 add reduce=True arg to HingeEmbeddingLoss
https://github.com/pytorch/pytorch/pull/5128 Fix ffi cdata for Variables.
https://github.com/pytorch/pytorch/pull/5127 Improve Variable interface
https://github.com/pytorch/pytorch/pull/5125 warn that CUDA capability 3.0 and 5.0 is no longer supported
https://github.com/pytorch/pytorch/pull/5119 Print Parameters like Variables (i.e. print scalars correctly).
https://github.com/pytorch/pytorch/pull/5117 Implement Variable.new(...) overloads for sparse tensors
https://github.com/pytorch/pytorch/pull/5115 Ensure Tensors have storages in resizeNd
https://github.com/pytorch/pytorch/pull/5114 Allow and warn when indexing a zero-dim Variable
https://github.com/pytorch/pytorch/pull/5113 Support calling pack_padedd_sequence with a Variable lengths
https://github.com/pytorch/pytorch/pull/5097 add reduce=True arg to MultiLabelSoftMarginLoss
https://github.com/pytorch/pytorch/pull/5093 Fix CPU torch.multinomial with noncontiguous prob tensor
https://github.com/pytorch/pytorch/pull/5090 Add Variable.item()
https://github.com/pytorch/pytorch/pull/5089 Check that indices and values are on the same device
https://github.com/pytorch/pytorch/pull/5088 Add deprecated add_out overload
https://github.com/pytorch/pytorch/pull/5085 Check shape instead of number of elements for some losses
https://github.com/pytorch/pytorch/pull/5080 Implement hinge_embedding_loss as a native function.
https://github.com/pytorch/pytorch/pull/5077 Restore torch.mm behavior for sparse variables
https://github.com/pytorch/pytorch/pull/5071  add reduce=True arg to SoftMarginLoss
https://github.com/pytorch/pytorch/pull/5064 DDP: 10% of NCCL backend perf improvements with mixed-prec support
https://github.com/pytorch/pytorch/pull/5054 Use fast integer division algorithm to avoid division ops inside kernels.
https://github.com/pytorch/pytorch/pull/5036 Use blocks machinery to simplify bookkeeping in autodiff
https://github.com/pytorch/pytorch/pull/5035 Bring back Tensor::data<__half>() and remove Tensor::data() template
https://github.com/pytorch/pytorch/pull/5030 Replace edge_type with Edge and create Variable::gradient_edge()
https://github.com/pytorch/pytorch/pull/5019 Add .clang-format
https://github.com/pytorch/pytorch/pull/5018 Remove FunctionFlags
https://github.com/pytorch/pytorch/pull/5017 Expose sparse variable sspaddmm
https://github.com/pytorch/pytorch/pull/5016 Expose sparse variable addmm, addmm_
https://github.com/pytorch/pytorch/pull/5010 add AVX2 implementation for sigmoid function
https://github.com/pytorch/pytorch/pull/5003 Don't allow scalars where vectors are required in mv, addmv, ger, addr.
https://github.com/pytorch/pytorch/pull/5000 Reverts force_gpu_half changes from #3660
https://github.com/pytorch/pytorch/pull/4999 Replace async with non_blocking for Python 3.7
https://github.com/pytorch/pytorch/pull/4992 Make cat/cat_out native function that rejects scalar inputs.
https://github.com/pytorch/pytorch/pull/4991 Only check that arguments are Variables in VariableType
https://github.com/pytorch/pytorch/pull/4982 Initial GraphExecutor Implementation.
https://github.com/pytorch/pytorch/pull/4980 Revert "Only check that arguments are Variables in VariableType (#4943)"
https://github.com/pytorch/pytorch/pull/4978 DDP: coalescing many little broadcasts to improve performance
https://github.com/pytorch/pytorch/pull/4977 Support stack_out as a native function.
https://github.com/pytorch/pytorch/pull/4975 Remove volatile section from autograd notes
https://github.com/pytorch/pytorch/pull/4972 Don't allow scalars in torch.dot for Variables.
https://github.com/pytorch/pytorch/pull/4967 Revert "torch.set_num_threads sets MKL option too"
https://github.com/pytorch/pytorch/pull/4966 Use TypeError in PythonArgParser
https://github.com/pytorch/pytorch/pull/4964 Operate on Variables in torch.nn.init
https://github.com/pytorch/pytorch/pull/4951 Properly fill in make_non_contiguous data for sizes that can't be mad…
https://github.com/pytorch/pytorch/pull/4949 torch.set_num_threads sets MKL option too
https://github.com/pytorch/pytorch/pull/4947 Initial type hints for function_wrapper
https://github.com/pytorch/pytorch/pull/4943 Only check that arguments are Variables in VariableType
https://github.com/pytorch/pytorch/pull/4933 fix copy/paste error in debug message in rnn.py
https://github.com/pytorch/pytorch/pull/4931 Add assignment support for Sequential
https://github.com/pytorch/pytorch/pull/4924 add reduce=True argument to MultiLabelMarginLoss
https://github.com/pytorch/pytorch/pull/4922 [ready] Layer Normalization
https://github.com/pytorch/pytorch/pull/4921 Release NCCL distributed backend from experimental
https://github.com/pytorch/pytorch/pull/4919 Don't use Variable vs. Tensor type-checks for requires_grad logic
https://github.com/pytorch/pytorch/pull/4911 Fixes to native_functions.yaml to match existing Tensor behavior
https://github.com/pytorch/pytorch/pull/4910 Add Linux Jenkins scripts to PyTorch repo.
https://github.com/pytorch/pytorch/pull/4909 Fix condition in inferUnsqueezeGeometry
https://github.com/pytorch/pytorch/pull/4892 Fix visibility of AT_CUDA_ENABLED
https://github.com/pytorch/pytorch/pull/4891 Added mixed-precision support in distributed training
https://github.com/pytorch/pytorch/pull/4889 Fix some scalar issues with autograd.
https://github.com/pytorch/pytorch/pull/4886 Deprecate out-of-place resize and resize_as on Variables.
https://github.com/pytorch/pytorch/pull/4883 Fix torch.pstrf on Variables
https://github.com/pytorch/pytorch/pull/4882 Implement sparse tensor and variable norm(value)
https://github.com/pytorch/pytorch/pull/4878 Make TensorDescriptor call more portable
https://github.com/pytorch/pytorch/pull/4876 Addition of ExponentialFamily
https://github.com/pytorch/pytorch/pull/4873 Added Poisson self KL + Bernoulli/Poisson KL
https://github.com/pytorch/pytorch/pull/4870 Slightly improve DistributedDataParallel (single-GPU binding) multi-process distributed training performance
https://github.com/pytorch/pytorch/pull/4861 Cuda 9.1 is cuda version 9010 not 9100
https://github.com/pytorch/pytorch/pull/4854 Fix deepcopy with scalars.
https://github.com/pytorch/pytorch/pull/4853 Various indexing fixes around scalars.
https://github.com/pytorch/pytorch/pull/4847 fix binary version scheme to be PEP compliant
https://github.com/pytorch/pytorch/pull/4826 Removed redundant import re
https://github.com/pytorch/pytorch/pull/4824 parallelize vol2col and col2vol of Conv3D with CPU backend
https://github.com/pytorch/pytorch/pull/4812 Fix output_nr not incremented correctly
https://github.com/pytorch/pytorch/pull/4811 Update pybind11
https://github.com/pytorch/pytorch/pull/4810 Create issue template with guidelines for issue submissions
https://github.com/pytorch/pytorch/pull/4807 Fix #4480 by tracing inputs before running function.
https://github.com/pytorch/pytorch/pull/4803 More efficient squeeze() backward in edge case
https://github.com/pytorch/pytorch/pull/4799 Add symbolic_override_first_arg_based
https://github.com/pytorch/pytorch/pull/4795 Enabling Infiniband support for Gloo data channel with auto IB detection
https://github.com/pytorch/pytorch/pull/4791 Favor Variables over Tensors for scalar constructors in torch.distrib…
https://github.com/pytorch/pytorch/pull/4788 Initialize cuda before setting cuda tensor types as default
https://github.com/pytorch/pytorch/pull/4787 Restore cuda variable.bernoulli()
https://github.com/pytorch/pytorch/pull/4786 Use Variable instead of Tensor in Function.forward
https://github.com/pytorch/pytorch/pull/4785 Restore sparse variable _dimI() and _dimV()
https://github.com/pytorch/pytorch/pull/4783 Fix squeeze() backward in edge case
https://github.com/pytorch/pytorch/pull/4780 Restore more sparse variable methods
https://github.com/pytorch/pytorch/pull/4779 Restore sparse variable transpose_() and t_()
https://github.com/pytorch/pytorch/pull/4775 New index computation strategy in Functions.cpp (Tensor/TensorList)
https://github.com/pytorch/pytorch/pull/4772 Use variadic templates instead of initializer lists and overloads (ROUND 2)
https://github.com/pytorch/pytorch/pull/4771 Implement Transforms
https://github.com/pytorch/pytorch/pull/4766 Removing NCCL clear_group_cache workaround with one more check in new_group
https://github.com/pytorch/pytorch/pull/4760 Temporary fix for Embedding fp16
https://github.com/pytorch/pytorch/pull/4754 Allow assertEqual checks with mixed Tensors, Variables, numbers.
https://github.com/pytorch/pytorch/pull/4753 Implement a (data-only) Variable factory
https://github.com/pytorch/pytorch/pull/4748 Add kwarg-only 'requires_grad' parameter to Variable factories.
https://github.com/pytorch/pytorch/pull/4746 Heuristic-based autograd execution order
https://github.com/pytorch/pytorch/pull/4745 Fix resize_as_ on Variables containing SparseTensors
https://github.com/pytorch/pytorch/pull/4744 Ensure that Tensors always have Storages
https://github.com/pytorch/pytorch/pull/4735 Legacy Padding: correct output size with nInputDim
https://github.com/pytorch/pytorch/pull/4734 Fused fixes
https://github.com/pytorch/pytorch/pull/4730 Allow Python Variables to be bound to at::Tensor in pybind11 converter
https://github.com/pytorch/pytorch/pull/4728 Implement record_stream on Variable
https://github.com/pytorch/pytorch/pull/4725 Allow Variables in calls to NCCL bindings.
https://github.com/pytorch/pytorch/pull/4724 Allow Variables in calls to type2backend
https://github.com/pytorch/pytorch/pull/4721 Define CHECK in torch/csrc/cuda/nccl.h
https://github.com/pytorch/pytorch/pull/4717 Make Symbol a true struct
https://github.com/pytorch/pytorch/pull/4712 Move repeat to torch/_utils.py
https://github.com/pytorch/pytorch/pull/4711 Replace PowConstant
https://github.com/pytorch/pytorch/pull/4707 Remove setting coalesce to 0 in sparse transpose_
https://github.com/pytorch/pytorch/pull/4705 adds reduce argument to BCEWithLogitsLoss interface
https://github.com/pytorch/pytorch/pull/4696 Add proper scalar checks to functions bound by nn.yaml.
https://github.com/pytorch/pytorch/pull/4691 Ensure lazy evaluation for probs and logits
https://github.com/pytorch/pytorch/pull/4690 Improve the engine support for functional graph execution
https://github.com/pytorch/pytorch/pull/4687 Restores some sparse variable methods
https://github.com/pytorch/pytorch/pull/4686 Fix embedding with sparse=True
https://github.com/pytorch/pytorch/pull/4683 Add print support for sparse variables
https://github.com/pytorch/pytorch/pull/4667 Local Response Normalization
https://github.com/pytorch/pytorch/pull/4663 replace full stop by comma
https://github.com/pytorch/pytorch/pull/4654 NLLLoss: current code works with dim = 3, so I added it to dim checks
https://github.com/pytorch/pytorch/pull/4643 Dataloader issues
https://github.com/pytorch/pytorch/pull/4640 Fixed non-determinate preprocessing on DataLoader
https://github.com/pytorch/pytorch/pull/4618 Adding is process_group initialized support
https://github.com/pytorch/pytorch/pull/4617 Add missing torch declarations to derivatives.yaml.
https://github.com/pytorch/pytorch/pull/4615 Implement MM fusion (MM with add reduction tree)
https://github.com/pytorch/pytorch/pull/4614 Allow broadcasting of value x params in Categorical
https://github.com/pytorch/pytorch/pull/4611 Fused fp16 lstm backward math fix.
https://github.com/pytorch/pytorch/pull/4604 Fix errors in travis config
https://github.com/pytorch/pytorch/pull/4598 spelling fix
https://github.com/pytorch/pytorch/pull/4586 Introduce a (non-public) autograd scalar method and improve printing
https://github.com/pytorch/pytorch/pull/4566 Remove accumulate_grad version_counter check.
https://github.com/pytorch/pytorch/pull/4565 Bind functions with out= arguments in VariableType
https://github.com/pytorch/pytorch/pull/4558 Further fix to tracing scope
https://github.com/pytorch/pytorch/pull/4545 Fix a missing AutoGPU
https://github.com/pytorch/pytorch/pull/4543 Ensure convolution weights are contiguous, fixes #4500
https://github.com/pytorch/pytorch/pull/4541 Link NNPACK even when CUDA is not available.
https://github.com/pytorch/pytorch/pull/4538 Fix torch.diag backward with non-square matrix
https://github.com/pytorch/pytorch/pull/4529 Fix return type for Bernoulli enumerate_support
https://github.com/pytorch/pytorch/pull/4527 Fix the inconsistency of `polygamma` on Tensor and Variable, for issue #4466
https://github.com/pytorch/pytorch/pull/4521 Fix abs specialization for `uint8_t` type.
https://github.com/pytorch/pytorch/pull/4512  Implement backward pass for pack_padded_sequence
https://github.com/pytorch/pytorch/pull/4511 Methods for checking CUDA memory usage
https://github.com/pytorch/pytorch/pull/4504 Fix multi-gpu fuser bug
https://github.com/pytorch/pytorch/pull/4496 Padding_idx in Embedding supports negative indexing
https://github.com/pytorch/pytorch/pull/4491 Add Slicing capabilities for Sequential, ModuleList and ParameterList
https://github.com/pytorch/pytorch/pull/4487 Refactor gen_variable_type
https://github.com/pytorch/pytorch/pull/4486 Fix handling of empty indices in CUDA Tensor.put_
https://github.com/pytorch/pytorch/pull/4473 Fix template type for std::array size
https://github.com/pytorch/pytorch/pull/4465 Rename native/TensorGeometry to native/TensorShape since there is alr…
https://github.com/pytorch/pytorch/pull/4462 Fix some scalar checks
https://github.com/pytorch/pytorch/pull/4460 Fix CUDA double backwards
https://github.com/pytorch/pytorch/pull/4448 Supporting logits as parameters in Bernoulli and Categorical
https://github.com/pytorch/pytorch/pull/4445 Remove assign_(Scalar).
https://github.com/pytorch/pytorch/pull/4430 [WIP] Added method cuda to PackedSequence.
https://github.com/pytorch/pytorch/pull/4421 Improve precision of dirichlet_grad() approximation
https://github.com/pytorch/pytorch/pull/4415 Modify derivatives for efficiency and change `destination` to `result` for consistency
https://github.com/pytorch/pytorch/pull/4414 fix distutils error for Unix case
https://github.com/pytorch/pytorch/pull/4409 RNN support has been implemented
https://github.com/pytorch/pytorch/pull/4401 Support NO_NNPACK environment variable
https://github.com/pytorch/pytorch/pull/4399 Add low-precision digamma() and polygamma() functions
https://github.com/pytorch/pytorch/pull/4398 Split cuda native functions into components; fix mistake with conv_tb…
https://github.com/pytorch/pytorch/pull/4395 Don't special case NN functions in gen_variable_type.py
https://github.com/pytorch/pytorch/pull/4393 Add Tensor::print() for gdb use.
https://github.com/pytorch/pytorch/pull/4389 Fix type signature of in-place NN functions
https://github.com/pytorch/pytorch/pull/4385 Update derivative of expm1
https://github.com/pytorch/pytorch/pull/4383 removes duplicate variable reference crash from pad_sequences
https://github.com/pytorch/pytorch/pull/4371 Adding description for Optimizers
https://github.com/pytorch/pytorch/pull/4370 Split off load_derivatives and gen_autograd_functions from gen_variable_type
https://github.com/pytorch/pytorch/pull/4369 Improve precision of standard_gamma_grad()
https://github.com/pytorch/pytorch/pull/4367 Fix creating tensors with np.longlong array
https://github.com/pytorch/pytorch/pull/4366 VariableType clean-up
https://github.com/pytorch/pytorch/pull/4350 Adding torch.expm1() and its inplace function
https://github.com/pytorch/pytorch/pull/4339 fix NameError in torch/nn/modules/rnn.py
https://github.com/pytorch/pytorch/pull/4331 Make expect file directory search more robust.
https://github.com/pytorch/pytorch/pull/4326 Reorder native_functions.yaml by alphabetical order.
https://github.com/pytorch/pytorch/pull/4325 Split NativeFunctions.cpp into functional components.
https://github.com/pytorch/pytorch/pull/4318 Fix btrifact for variables
https://github.com/pytorch/pytorch/pull/4314 fix AMSGrad for SparseAdam
https://github.com/pytorch/pytorch/pull/4312 Vectorize normal_
https://github.com/pytorch/pytorch/pull/4308 Make Variable.is_sparse an attribute
https://github.com/pytorch/pytorch/pull/4307 Fix default device for Variable.new()
https://github.com/pytorch/pytorch/pull/4306 Throw exception in checkBackend, improve standard_gamma_grad error me…
https://github.com/pytorch/pytorch/pull/4303 Add factory Type::sparse_coo_tensor(indices, values)
https://github.com/pytorch/pytorch/pull/4301 Use `where` rather than `_s_where` in `_s_where` backwards so `where`…
https://github.com/pytorch/pytorch/pull/4298 Enable functional torch.where.
https://github.com/pytorch/pytorch/pull/4267 Remove unused thnn/loss.py
https://github.com/pytorch/pytorch/pull/4262 Ensure gamma samples are positive
https://github.com/pytorch/pytorch/pull/4259 Implement torch.where(condition, x, y) CPU Variable.
https://github.com/pytorch/pytorch/pull/4255 Remove template_scalar, implement is_signed using dispatch.
https://github.com/pytorch/pytorch/pull/4249 Don't mark index as traceable, and other improvements
https://github.com/pytorch/pytorch/pull/4244 Further relax VariableFlags, ... and fix bugs
https://github.com/pytorch/pytorch/pull/4242 Translate None to zeros for old-style autograd functions
https://github.com/pytorch/pytorch/pull/4238 Deprecate nn.NLLLoss2d
https://github.com/pytorch/pytorch/pull/4231 Add reduce arg to BCELoss
https://github.com/pytorch/pytorch/pull/4221 Revert "Add reduce arg to BCELoss"
https://github.com/pytorch/pytorch/pull/4209 Make import work even if 'tools' is available in Python path
https://github.com/pytorch/pytorch/pull/4200 Expose node scopeName to python
https://github.com/pytorch/pytorch/pull/4191 Relax verify of VariableFlags
https://github.com/pytorch/pytorch/pull/4185 Fix another leak in pybind11 code.
https://github.com/pytorch/pytorch/pull/4184 Preprocess both inplace and non-inplace nn functions
https://github.com/pytorch/pytorch/pull/4183 Allowing usage of GPU Direct within PyTorch for the Broadcast operation
https://github.com/pytorch/pytorch/pull/4182 Fix a bug where from_dlpack fails if cuda is not initialized.
https://github.com/pytorch/pytorch/pull/4174 Rearrange dimensions for pointwise operations for better performance.
https://github.com/pytorch/pytorch/pull/4163 support RNN export
https://github.com/pytorch/pytorch/pull/4158 Re-initialize autograd engine in child processes
https://github.com/pytorch/pytorch/pull/4143 Improve symbolic hack a bit
https://github.com/pytorch/pytorch/pull/4142 add reparameterization, combine sample and sample_n
https://github.com/pytorch/pytorch/pull/4135 Add an option to suppress download progress
https://github.com/pytorch/pytorch/pull/4124 Implement Variable._sparse_mask
https://github.com/pytorch/pytorch/pull/4113 Add size checks for sparse tensor constructor
https://github.com/pytorch/pytorch/pull/4096 Refactor generation of NN derivatives
https://github.com/pytorch/pytorch/pull/4095 Add python only default init expression; Implement stft, hann/hamming/bartlett window.
https://github.com/pytorch/pytorch/pull/4094 Implement pin_memory() as a NativeFunction
https://github.com/pytorch/pytorch/pull/4091 Enable half communication for distributed
https://github.com/pytorch/pytorch/pull/4089 Allow specification of bool defaults in native functions.
https://github.com/pytorch/pytorch/pull/4088 Expose resize_ and resize_as_ to Python
https://github.com/pytorch/pytorch/pull/4082 Implement Variable.__invert__
https://github.com/pytorch/pytorch/pull/4080 Implement Variable.new
https://github.com/pytorch/pytorch/pull/4079 Add missing derivatives.yaml input
https://github.com/pytorch/pytorch/pull/4075 Implement neg for all types
https://github.com/pytorch/pytorch/pull/4074 Ensure RNNCell variants don't broadcast
https://github.com/pytorch/pytorch/pull/4063 Fix non-determinism in code generation scripts
https://github.com/pytorch/pytorch/pull/4062 Allow .view on noncontig tensors when certain conditions are met
https://github.com/pytorch/pytorch/pull/4061 Adding index_select to symbolic.py
https://github.com/pytorch/pytorch/pull/4059 throw new -> throw
https://github.com/pytorch/pytorch/pull/4057 Implement apply_, map_, and map2_ in Variable
https://github.com/pytorch/pytorch/pull/4045 Add Variable._cdata
https://github.com/pytorch/pytorch/pull/4044 handle requires_grad when creating buckets for distributed
https://github.com/pytorch/pytorch/pull/4043 Implement Variable.from_numpy
https://github.com/pytorch/pytorch/pull/4042 Enable OpenMP in fuser
https://github.com/pytorch/pytorch/pull/4038 Implement Variable.tolist()
https://github.com/pytorch/pytorch/pull/4036 Implement Variable.storage_type()
https://github.com/pytorch/pytorch/pull/4034 added AMSgrad optimizer to Adam and SparseAdam
https://github.com/pytorch/pytorch/pull/4026 Add debug and fix some bugs in CPU fuser
https://github.com/pytorch/pytorch/pull/4020 Delete _write_metadata and move _new_with_metadata_file into Python
https://github.com/pytorch/pytorch/pull/4016 Fix the symbolic for view
https://github.com/pytorch/pytorch/pull/4015 Add is_pinned, is_shared, and share_memory_ to Variable
https://github.com/pytorch/pytorch/pull/4011 Exclude attrs with invalid python variable names from __dir__
https://github.com/pytorch/pytorch/pull/4006 Implement Variable.numpy()
https://github.com/pytorch/pytorch/pull/3978 Implement reparameterized gradient for Gamma sampler
https://github.com/pytorch/pytorch/pull/3972 Implements gradients calculation for trtrs
https://github.com/pytorch/pytorch/pull/3970 Replace Variable.volatile with torch.no_grad()
https://github.com/pytorch/pytorch/pull/3968 Add streams and comms as optional arguments
https://github.com/pytorch/pytorch/pull/3961 Add a CPU Fuser (single core)
https://github.com/pytorch/pytorch/pull/3952 Update CONTRIBUTING.md
https://github.com/pytorch/pytorch/pull/3943 Implement matmul as a native function; use it for Variable impl
https://github.com/pytorch/pytorch/pull/3932 use torch.cat in _flatten
https://github.com/pytorch/pytorch/pull/3926 Fix indexing with all zero ByteTensors
https://github.com/pytorch/pytorch/pull/3907 Remove Function::is_executable
https://github.com/pytorch/pytorch/pull/3903 Fix _analytical_jacobian to not require in-place grad accumulation
https://github.com/pytorch/pytorch/pull/3888 Clean up InputBuffer
https://github.com/pytorch/pytorch/pull/3881 Correct gradient of rosenbrock
https://github.com/pytorch/pytorch/pull/3875 Pad sequences and Pack sequences
https://github.com/pytorch/pytorch/pull/3866 Add interpreter support for Handles/PythonOp/CppOp
https://github.com/pytorch/pytorch/pull/3862 Set seed at top-level of common.py
https://github.com/pytorch/pytorch/pull/3856 Delete unused autograd functions
https://github.com/pytorch/pytorch/pull/3846 Use warp shuffles in cuda varInnermostDim
https://github.com/pytorch/pytorch/pull/3843 Fix CharType min and max values
https://github.com/pytorch/pytorch/pull/3838 Implement is_sparse/is_distributed as native functions, have cpu() not change density
https://github.com/pytorch/pytorch/pull/3837 - added size_splits to functional
https://github.com/pytorch/pytorch/pull/3831 Fix errors in previous DataChannelMPI refactor
https://github.com/pytorch/pytorch/pull/3820 Make integer parameters and buffers immune to float(), double() and half()
https://github.com/pytorch/pytorch/pull/3817 Improve DataChannelMPI
https://github.com/pytorch/pytorch/pull/3816 Add determinant function on variable; Add backward on svd
https://github.com/pytorch/pytorch/pull/3795 Reflect renaming of OS X to macOS
https://github.com/pytorch/pytorch/pull/3771 fix elapsed_us spelling
https://github.com/pytorch/pytorch/pull/3765 Implement Variable.storage()
https://github.com/pytorch/pytorch/pull/3754 CUDA mode profiler fixes
https://github.com/pytorch/pytorch/pull/3750 Add Tensor.slice()
https://github.com/pytorch/pytorch/pull/3734 Add cudaEvent support to the profiler
https://github.com/pytorch/pytorch/pull/3733 Fix symbolic for Embedding and Upsampling and improve error messages
https://github.com/pytorch/pytorch/pull/3727 Record stack traces for CppOp's
https://github.com/pytorch/pytorch/pull/3707 Implement VariableType::alias
https://github.com/pytorch/pytorch/pull/3706 Implement toBackend and toScalarType on VariableType
https://github.com/pytorch/pytorch/pull/3700 fix half uniform for cuda 7.5
https://github.com/pytorch/pytorch/pull/3692 set CC and CXX only when it's empty
https://github.com/pytorch/pytorch/pull/3683 Split off in-place NN functions
https://github.com/pytorch/pytorch/pull/3681 Implement bmm symbolic
https://github.com/pytorch/pytorch/pull/3680 Propagate is_volatile to the base when performing in-place ops on views
https://github.com/pytorch/pytorch/pull/3679 Fix a reference cycle when in-place ops on views save the output
https://github.com/pytorch/pytorch/pull/3676 Move detach to variable
https://github.com/pytorch/pytorch/pull/3672 NativeFunctions: support backend-specific dispatch and SpatialRoIPooling
https://github.com/pytorch/pytorch/pull/3658 Cast tensors when loading optimizer state dicts
https://github.com/pytorch/pytorch/pull/3656 Solved boolean ambiguity for variables and tensors which contain one value.
https://github.com/pytorch/pytorch/pull/3629 fix for compilation problems with PRId64 format specifier
https://github.com/pytorch/pytorch/pull/3616 Allow 1->N broadcasts at the beginning and end to be fused
https://github.com/pytorch/pytorch/pull/3609 Ensure that Variables are at least one-dim in VariableType
https://github.com/pytorch/pytorch/pull/3602 Improvements around torch.cat on empty Variables
https://github.com/pytorch/pytorch/pull/3549 Previous PyTorch version info
https://github.com/pytorch/pytorch/pull/3532 Add reduce arg to BCELoss
https://github.com/pytorch/pytorch/pull/3526 Reuse intermediate results over multiple backwards grad_inputs
https://github.com/pytorch/pytorch/pull/3520 adds flag __CUDA_NO_HALF_OPERATORS__
https://github.com/pytorch/pytorch/pull/3509 Optimizer: optimize transposes in variety of circumstances
https://github.com/pytorch/pytorch/pull/3505 improvements to ModuleList and ParameterList classes
https://github.com/pytorch/pytorch/pull/3501 implement `__dir__`for Variable
https://github.com/pytorch/pytorch/pull/3500 ignore digit in container's `__dir__`
https://github.com/pytorch/pytorch/pull/3480 Two miscellaneous fixes
https://github.com/pytorch/pytorch/pull/3465 Generate native functions with const ref Tensor arguments.
https://github.com/pytorch/pytorch/pull/3444 added missing arg and improved example clarity
https://github.com/pytorch/pytorch/pull/3435 Implemented NCCL Distributed Backend for PyTorch with new dist APIs
https://github.com/pytorch/pytorch/pull/3429 Allow empty index tensor for index_select
https://github.com/pytorch/pytorch/pull/3422 Register VariableType calls in autograd profiler
https://github.com/pytorch/pytorch/pull/3411 Clear out eigenvector tensor when eigenvector=F for symeig
https://github.com/pytorch/pytorch/pull/3409 Add Tensor Core ops to RNNs for Volta
https://github.com/pytorch/pytorch/pull/3408 lazy-load nvrtc and libcuda
https://github.com/pytorch/pytorch/pull/3384 Allow in-place operations on views
https://github.com/pytorch/pytorch/pull/3382 Implement reduce keyword for SmoothL1Loss
https://github.com/pytorch/pytorch/pull/3371 Pretty names: support names set via export or Variable constructor
https://github.com/pytorch/pytorch/pull/3370 Follow up #3211 (sparse broadcast_coalesced, reduce_add_coalesced)
https://github.com/pytorch/pytorch/pull/3366 Add reduce keyword to L1Loss
https://github.com/pytorch/pytorch/pull/3352 fix some more mathjax
https://github.com/pytorch/pytorch/pull/3341 add gumbel_softmax, based on Eric Jang's implementation
https://github.com/pytorch/pytorch/pull/3336 Prevent numerical issues with poisson_nll_loss when log_input=False
https://github.com/pytorch/pytorch/pull/3263 Add torch.take and Tensor.put_
https://github.com/pytorch/pytorch/pull/3127 adaptive pooling supports only specifying size in certain dimension
https://github.com/pytorch/pytorch/pull/3016 Introduce scopes during tracing
https://github.com/pytorch/pytorch/pull/2953 WIP: add numpy() and from_numpy() to HalfTensor
https://github.com/pytorch/pytorch/pull/2764 [Done]parallelize elementwise operation with openmp