Soumith Chintala soumith
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| https://github.com/pytorch/pytorch/pull/3526 Reuse intermediate results over multiple backwards grad_inputs | |
| https://github.com/pytorch/pytorch/pull/3509 Optimizer: optimize transposes in variety of circumstances | |
| https://github.com/pytorch/pytorch/pull/3409 Add Tensor Core ops to RNNs for Volta | |
| https://github.com/pytorch/pytorch/pull/3370 Follow up #3211 (sparse broadcast_coalesced, reduce_add_coalesced) | |
| https://github.com/pytorch/pytorch/pull/3336 Prevent numerical issues with poisson_nll_loss when log_input=False | |
| https://github.com/pytorch/pytorch/pull/2764 [Done]parallelize elementwise operation with openmp | |
| https://github.com/pytorch/pytorch/pull/6110 Fix bilinear performance regression | |
| https://github.com/pytorch/pytorch/pull/6078 Exp, log, sin, cos vectorized | |
| https://github.com/pytorch/pytorch/pull/6062 Enable MKLDNN convolution forward and backward | |
| https://github.com/pytorch/pytorch/pull/6026 Speed up sum over a dimension |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| https://github.com/pytorch/pytorch/pull/6553 Restore allow_unused functionality | |
| https://github.com/pytorch/pytorch/pull/6173 Update torch.nn.init and torch.nn.utils.clip_grad |
soumith
/ feature.txt
Created
April 23, 2018 14:28
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| 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/3435 Implemented NCCL Distributed Backend for PyTorch with new dist APIs | |
| https://github.com/pytorch/pytorch/pull/3384 Allow in-place operations on views | |
| https://github.com/pytorch/pytorch/pull/3341 add gumbel_softmax, based on Eric Jang's implementation | |
| https://github.com/pytorch/pytorch/pull/3263 Add torch.take and Tensor.put_ | |
| https://github.com/pytorch/pytorch/pull/2953 WIP: add numpy() and from_numpy() to HalfTensor | |
| https://github.com/pytorch/pytorch/pull/6528 Support arbitrary number of batch dimensions in *FFT | |
| 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 |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| https://github.com/pytorch/pytorch/pull/3429 Allow empty index tensor for index_select | |
| https://github.com/pytorch/pytorch/pull/3411 Clear out eigenvector tensor when eigenvector=F for symeig | |
| https://github.com/pytorch/pytorch/pull/3127 adaptive pooling supports only specifying size in certain dimension | |
| https://github.com/pytorch/pytorch/pull/6541 Fix regression that STFT has no backward. | |
| https://github.com/pytorch/pytorch/pull/6517 More precise digamma | |
| 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/6418 [pytorch] Fix clamp is missing kwarg out (#6028) | |
| 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 |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| 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 |
soumith
/ 0.4_onnx_pr
Created
April 18, 2018 18:52
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| https://github.com/pytorch/pytorch/pull/6651 Export onnx protobuf bindings to python | |
| https://github.com/pytorch/pytorch/pull/6574 Follow the change of ONNX Cast operator "to" attribute | |
| https://github.com/pytorch/pytorch/pull/6560 Create CODEOWNERS entry for torch/onnx | |
| https://github.com/pytorch/pytorch/pull/6331 Fix the c2-onnx exporter bug on Gemm | |
| https://github.com/pytorch/pytorch/pull/6243 fix broadcast export to onnx | |
| https://github.com/pytorch/pytorch/pull/6220 Support export torch.max(input, dim) and torch.min(input, dim) to ONNX | |
| https://github.com/pytorch/pytorch/pull/6101 Fix a bug in ONNX symbolic of average 3d pooling op | |
| https://github.com/pytorch/pytorch/pull/6016 Experimental support for different ONNX export types | |
| https://github.com/pytorch/pytorch/pull/5972 Update no_unions flag for nanopb gen and update ONNX proto files | |
| https://github.com/pytorch/pytorch/pull/5935 Specify outputs number in embedding_bag onnx export |
soumith
/ 0.4_distributions_pr
Created
April 18, 2018 18:51
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| https://github.com/pytorch/pytorch/pull/6666 [distributions] Skip validation of lazy properties | |
| https://github.com/pytorch/pytorch/pull/6615 [distributions] Implement Independent distribution | |
| https://github.com/pytorch/pytorch/pull/6427 Add method to calculate perplexity of distribution | |
| https://github.com/pytorch/pytorch/pull/6172 [distributions] KL-Divergence for Multivariate Normal | |
| https://github.com/pytorch/pytorch/pull/6035 [distributions] Support python floats in AffineTransform | |
| https://github.com/pytorch/pytorch/pull/5989 [distributions] Rename .params to .arg_constraints, fix logic | |
| https://github.com/pytorch/pytorch/pull/5976 [distributions] Implement Power transform | |
| https://github.com/pytorch/pytorch/pull/5931 [distributions] Fix scalar bugs in torch.distributions.transforms etc. | |
| https://github.com/pytorch/pytorch/pull/5910 [distributions] Support pickling of constraint objects | |
| https://github.com/pytorch/pytorch/pull/5842 [distributions] Avoid in-place ops in BoltzmannTransform |
soumith
/ Dockerfile
Created
April 3, 2018 16:47
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| FROM nvidia/cuda:8.0-devel-centos7 | |
| ENV LC_ALL en_US.UTF-8 | |
| ENV LANG en_US.UTF-8 | |
| ENV LANGUAGE en_US.UTF-8 | |
| RUN yum install -y wget curl perl cmake util-linux xz bzip2 git | |
| RUN yum install -y patch | |
| RUN yum install -y yum-utils centos-release-scl |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| import torch | |
| import torch.nn as nn | |
| from torch.autograd import Variable | |
| torch.manual_seed(1) | |
| # do gradcheck | |
| N = 1 | |
| C = 1 | |
| D = 1 | |
| H = 2 | |
| W = 1 |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Scopes 0.3.1 backport https://github.com/pytorch/pytorch/pull/5153 | |
| Cherry pick dataloader issue fix to 0.3.1 https://github.com/pytorch/pytorch/pull/5140 | |
| Fixed double memory accesses of several pointwise operations. https://github.com/pytorch/pytorch/pull/5068 | |
| Broadcast output requires_grad only if corresponding input requires_grad https://github.com/pytorch/pytorch/pull/5061 | |
| Fix topk work size computation https://github.com/pytorch/pytorch/pull/5053 | |
| Fix maxpool3d / avgpool3d crashs https://github.com/pytorch/pytorch/pull/5052 | |
| Fix blas addmm (gemm) condition check https://github.com/pytorch/pytorch/pull/5048 | |
| Fix C FFI extension after moving TH to C++ https://github.com/pytorch/pytorch/pull/5005 | |
| make torch.set_num_threads also set MKL threads (take 2) https://github.com/pytorch/pytorch/pull/5002 | |
| Fix reduction functions to respect the stride of the output https://github.com/pytorch/pytorch/pull/4995 |