strisunshinewentingwang/pytorch

Pull Requests Commits

[ROCm] fix cub hipify mappings (#44431)

jeffdaily committed 4 years ago

65d4a6b7

[NNC] Make it able to normalize loop with variable start (#44133)

Cheng Chang committed 4 years ago

28bd4929

Add reset_grad() function (#44423)

taiyuanz committed 4 years ago

c5158811

[caffe2] Speed up compilation of aten-op.cc (#44440)

ot committed 4 years ago

6324ef4c

[JIT] Propagate type sharing setting to submodule compilation (#44226)

Meghan Lele committed 4 years ago

89ac30af

[JIT] Add support for del to TS classes (#44352)

Meghan Lele committed 4 years ago

d3b6d5ca

Expose the interface of nesterov of SGD Optimizer from caffe2 to dper

Gang Shen committed 4 years ago

058d7228

[caffe2] exposes Net cancellation through pybind state (#44043)

dahsh committed 4 years ago

5ee31308

Fix HashStoreTests and move to Gtest (#43384)

osalpekar committed 4 years ago

e028ad07

Modularize FileStoreTest and move to Gtest (#43383)

osalpekar committed 4 years ago

69a3ff00

Convert StoreTestUtils to Gtest (#43382)

osalpekar committed 4 years ago

a7fba7de

Improving ModuleList indexing error msg (#43361)

Elias Ellison committed 4 years ago

b69c28d0

use non-overflowing divide in cuda kernel util GET_BLOCKS (#44391)

Rong Rong committed 4 years ago

c010ef7f

[pyper] export caffe2 bucketize GPU operator to pytorch

xw285cornell committed 4 years ago

ba6ddaf0

Revert D23568330: [pytorch][PR] Moves some of TestTorchMathOps to OpInfos

Elias Ellison committed 4 years ago

e0c65abd

Small fixes in Dependency.cmake and run_test.py (#44414)

malfet committed 4 years ago

fc51047a

[JIT] Support partially specified sizes/strides in IRParser (#44113)

Lilyjjo committed 4 years ago

b0bcdbb1

[quant] quantized path for ConstantPadNd (#43304)

z-a-f committed 4 years ago

36742649

fix typo in embedding_bag_non_contiguous_weight test (#44382)

lixinyu committed 4 years ago

032480d3

[PyTorch][Mobile] Insert the module name as `name()` to metadata dict if metadata doesn't contain "model_name" (#44400)

husthyc committed 4 years ago

a00d36b0

Check commutativity for computed dispatch table and add a test to check entries. (#44088)

Ailing Zhang committed 4 years ago

24efd29d

[NCCL] Add Environment Variable to guard Async Error Handling feature (#44163)

osalpekar committed 4 years ago

48c47db8

[NCCL] ProcessGroupNCCL Destructor Blocks on WorkNCCL Completion (#41054)

osalpekar committed 4 years ago

211ece72

[NCCL] WorkNCCL Helper Functions (#41053)

osalpekar committed 4 years ago

afbf2f14

[NCCL] Abort Errored and Timed Out NCCL Communicators from Watchdog Thread (#41052)

osalpekar committed 4 years ago

f8f7b784

[NCCL] Use cudaEventQuery to Poll for GPU operation errors (#41051)

osalpekar committed 4 years ago

4e5c55ef

[NCCL] Timeout Loop Thread for Async Error Handling (#41050)

osalpekar committed 4 years ago

1df24fd4

Preserve .ninja_log in build artifacts (#44390)

malfet committed 4 years ago

15cbd1cf

[Reland] Optimize code path for adaptive_avg_pool2d when output size is (1, 1) (#44211)

xwang233 committed 4 years ago

ef4475f9

Benchmarks: make fuser and executor configurable from command line. (#44291)

Mikhail Zolotukhin committed 4 years ago

37093f4d

Newer Older