strisunshinewentingwang/pytorch

Pull Requests Commits

Revert D29135358: [quant] Input-Weight Equaliaztion - convert modifications

Rong Rong (AI Infra) committed 3 years ago

e60f9cfc

Fix parallel tbb build (#60532)

peterbell10 committed 3 years ago

03ab5b72

Add `NoChunk` wrapper for pipeline args. (#57325)

pritamdamania committed 3 years ago

bea83e2e

Use JOB_BASE_NAME throughout code--consolidate CIRCLE_JOB (#60425)

janeyx99 committed 3 years ago

63856210

Disable group group backend rpc tests from running on CI (#60407)

H-Huang committed 3 years ago

ff3678ee

Support non-Tensor args in the Pipe API (#57226)

pritamdamania committed 3 years ago

109f8314

Reland D29190420: [nnc][tests] Tests and benchmarks for computeSum (#60550)

bertmaher committed 3 years ago

10e11dbd

Port `any` kernel to structured kernels. (#60361)

ysiraichi committed 3 years ago

5fd45b80

Port `all` kernel to structured kernels. (#60360)

ysiraichi committed 3 years ago

a5aa940f

Fix convolution_depthwise3x3_winograd for multichannel output (#60460)

malfet committed 3 years ago

7b2d3751

Adding windows CUDA smoke tests on PRs (#59686)

janeyx99 committed 3 years ago

c63a0d0c

[DDP] Remove python GradBucket construction (#60301)

rohan-varma committed 3 years ago

8162439c

To add Nesterov Adam Algorithm to Optimizers (#59009)

iramazanli committed 3 years ago

e8690dac

Remove unused sample input argument from functions to resolve issue #55737 (#60486)

NivekT committed 3 years ago

a2525b03

Add device runtime API for the plug-in to register platform python module into torch (#59857)

chengjunlu committed 3 years ago

265f0e53

Fix test failures with some glibc libraries (#60450)

Flamefire committed 3 years ago

c97d4d5a

Clean Up ZeRO (#60285)

awgu committed 3 years ago

f0e4e4be

Ensure proper syncs for out-of-place grad creation (torch.autograd.grad) when backward ops run on side streams (#60127)

mcarilli committed 3 years ago

56481f97

Revert D29190420: [nnc][tests] Tests and benchmarks for computeSum

anjali411 committed 3 years ago

b14f19b6

To add edge_order=2 and documentation for gradient operator (#58165)

iramazanli committed 3 years ago

90cd57ee

[NormalizeArgs] Retain node.meta (#60449)

jfix71 committed 3 years ago

7ed07e2a

Ensure num_threads is initialized before calling omp_get_max_threads (#60185)

peterbell10 committed 3 years ago

66452e0a

OpenMP: Refactor parallel_reduce to share code with parallel_for (#60184)

peterbell10 committed 3 years ago

19553438

Ensure thread id is valid in nested parallel regions (#60183)

peterbell10 committed 3 years ago

c75714e5

Migrate crossKernel from THC to ATen (CUDA) (#60039)

peterbell10 committed 3 years ago

3f3fd570

[BE] Fix Convolution.cpp build warnings (#60463)

malfet committed 3 years ago

f590ccea

Increase tolerance for test_grad_scaling_clipping (#60458)

Flamefire committed 3 years ago

3846cef2

`topk` on CUDA supports `bfloat16` (#59977)

eqy committed 3 years ago

40de03fc

[nnc][tests] Tests and benchmarks for computeSum (#60160)

bertmaher committed 3 years ago

21479ad2

[nnc] Speed up batchnorm benchmark

bertmaher committed 3 years ago

fbeb8b49

Newer Older