Create your own GitHub profile
Sign up for your own profile on GitHub, the best place to host code, manage projects, and build software alongside 50 million developers.
Sign up
- New York, NY
- Sign in to view email
- http://ezyang.com
Pinned
3,177 contributions in the last year
Contribution activity
June 2020
Created a pull request in pytorch/pytorch that received 9 comments
Restore docs coverage test via sphinx (#38244)
Summary:
Fixes gh-37590
Adds an extra make coverage to document building, which uses the built-in facility in sphinx to check docstring coverage. A…
+92
−124
•
9
comments
- Add batching rule for torch.sum(tensor, dims)
- Changes TensorIterator computation to not consider out kwarg, lets UnaryOps safe cast to out
- Cleanup debug info switches with MSVC
- Add aarch64 ci badge
- Include AT_PARALLEL_OPENMP/AT_PARALLEL_NATIVE/AT_PARALLEL_NATIVE_TBB to ATen/Config.h
- Assert that kernels are called with the right signature
- Add experimental deterministic flag
- Fix smoke test jobs
- [ROCm] Enable AMP autocast tests on ROCm
- Show which type was the wrong one when a signature is invalid
- Impose maximum level restriction for BatchedTensors
- Fix IDE hint channels_last & preserve_format
- [pytorch] add operator name to callBoxed() error message
- [pytorch] remove tracing logic from gen_variable_factories.py
- Fix overflow issues when unpacking large numbers
- .circleci: Move binary builds into their own workflow
- [pytorch] move tracing logic to a separate dispatch backend
- Remove cuda init patch
- Add rocm image to circleci docker builder
- [resubmit] Use TensorMethods.cpp
- More fixes about using Windows API through ctypes
- Leak safety in RReLU
- Add "batching rule" for torch.sum(tensor, dims)
- `verbose` unused in `torch.backends.cudnn`
- Fix missing code in 'Installing C++ distribution of Pytorch'
- Some pull request reviews not shown.
Created an issue in pytorch/pytorch that received 2 comments
friend constexpr in templated struct loses constexpr-ness in nvcc
This fails to compile in nvcc:
template <typename T>
struct A { constexpr A(T x) : x_(x) {} constexpr bool equals(const A& other) { return x_ == ot…
2
comments
- Force JIT to do type inference even when mypy annotated
- test_nn_module_tests should run less tests
- [JIT] Cannot directly run tests in test_module_interface
- pytest suppresses stderr from Python startup by default
- Misannotation of layer_norm parameters causes internal assert failure
- TestTorchDeviceTypeCPU.test_complex_flip_cpu_complex128 - TypeError: list indices must be integers or slices, not tuple
- JIT test suite has dependencies across tests