strisunshinewentingwang/pytorch

Pull Requests Commits

[NNC] Registerizer for GPU [1/x] (#42606)

nickgg committed 4 years ago

aabdef51

align qlinear benchmark to linear benchmark (#42767)

vkuzo committed 4 years ago

57b056b5

align qconv benchmark to conv benchmark (#42761)

vkuzo committed 4 years ago

a7bdf575

Fix orgqr input size conditions (#42825)

kurtamohler committed 4 years ago

2c8cbd78

Introduce experimental FX library (#42741)

James Reed committed 4 years ago

575e7497

Modify clang code coverage to CMakeList.txt (for MacOS) (#42837)

scintiller committed 4 years ago

7524699d

Update the documentation for scatter to include streams parameter. (#42814)

jgulian committed 4 years ago

42114a01

Fix a typo in EmbeddingBag.cu (#42742)

xwang233 committed 4 years ago

1041bdeb

[JIT] Fix typing.Final for python 3.8 (#39568)

zasdfgbnm committed 4 years ago

91623528

Optimization with Backward Implementation of Learnable Fake Quantize Per Channel Kernel (CPU and GPU) (#42810)

Paul Shao committed 4 years ago

d28639a0

Raise error if `at::native::embedding` is given 0-D weight (#42550)

kurtamohler committed 4 years ago

42b4a713

Added torch::cuda::manual_seed(_all) to mirror torch.cuda.manual_seed(_all) (#42638)

heitorschueroff committed 4 years ago

d396d135

vmap: temporarily disable support for random functions (#42617)

zou3519 committed 4 years ago

e8f4b04d

Don't materialize output grads (#41821)

heitorschueroff committed 4 years ago

ffc3da35

Revert D23002043: add net transforms for fusion

Mike Ruberry committed 4 years ago

ddcf3ded

[quant] Sorting the list of dispathes (#42758)

z-a-f committed 4 years ago

59b10f79

Fix ROCm CI by increasing test timeout (#42827)

Mike Ruberry committed 4 years ago

dedcc30c

add net transforms for fusion (#42763)

Hector Yuen committed 4 years ago

a4b763bc

Fix "non-negative integer" error messages (#42734)

NTT123 committed 4 years ago

10388789

adaptive_avg_pool[23]d: check output_size.size() (#42831)

malfet committed 4 years ago

c14a7f68

[c10d] Template computeLengthsAndOffsets() (#42706)

jiayisuse committed 4 years ago

c9e82564

Skip test_c10d.ProcessGroupNCCLTest under TSAN (#42750)

rohan-varma committed 4 years ago

a414bd69

Rename some BatchedTensorImpl APIs (#42700)

zou3519 committed 4 years ago

a2559652

BatchedTensor fallback: extended to support ops with multiple Tensor returns (#42628)

zou3519 committed 4 years ago

8f67c7a6

test_cpp_rpc: Build test_e2e_process_group.cpp only if USE_GLOO is true (#42836)

malfet committed 4 years ago

64a7939e

[vulkan] cat op (concatenate) (#41434)

IvanKobzarev committed 4 years ago

87185245

Fix `torch.nn.functional.grid_sample` crashes if `grid` has NaNs (#42703)

malfet committed 4 years ago

3cf2551f

change pt_defs.bzl to python file (#42725)

linbinyu committed 4 years ago

e06b4be5

DDP communication hook: skip dividing grads by world_size if hook registered. (#42400)

sinannasir committed 4 years ago

752f433a

.circleci: Only do comparisons when available (#42816)

seemethere committed 4 years ago

d7aaa332

Newer Older