strisunshinewentingwang/pytorch

Pull Requests Commits

Add `torch.cuda.streams.ExternalStream` (#57781)

Emilio Castillo committed 3 years ago

d7ef9b73

Fix MaxPool default pad documentation (#59404)

Varal7 committed 3 years ago

c7693003

Fix broken hyperlinks (#59425)

awgu committed 3 years ago

6d51a897

Search for static OpenBLAS compiled with OpenMP (#59428)

malfet committed 3 years ago

63956610

.circleci: Disable USE_GOLD_LINKER for CUDA 10.2 (#59413)

seemethere committed 3 years ago

c7a3a13b

Merge TensorPipe's CPU and CUDA channel registry (#59375)

lw committed 3 years ago

06ed6583

Remove LazyStreamContext (2 out of 2) (#59299)

lw committed 3 years ago

c09beaaf

Remove LazyStreamContext (1 out of 2) (#59298)

lw committed 3 years ago

03a5c6ea

Fix CUDA sync when switching streams in RPC tests (#59297)

lw committed 3 years ago

3e7396f9

Fix race condition in TP agent (#58753)

lw committed 3 years ago

8f4cfaa9

Ensure async_execution works with CUDAFuture (#56863)

lw committed 3 years ago

c0acffa6

Avoid re-doing CUDA stream sync in OwnerRRef (#57355)

lw committed 3 years ago

7bcd8f94

[RPC Framework] Separate initialize_from_module_rref method out of RemoteModule constructor (#59292)

Yi Wang committed 3 years ago

d009c9c1

Fix symbolic derivative of hardswish (#59405)

bertmaher committed 3 years ago

c3bf42e0

[nnc] Add hardsigmoid (#59069)

bertmaher committed 3 years ago

9ac95478

[NNC] Add python bindings for Compute2 (#59350)

huiguoo committed 3 years ago

c717ce67

Make JIT not assume that the device is CUDA. (#54238)

chengjunlu committed 3 years ago

db90533b

[NNC] Fix loopnest.cache_accesses for reduce ops (fixed #59002) (#59136)

huiguoo committed 3 years ago

7c4ac9e3

[torch] Remove migration warning for ScriptDict

Meghan Lele committed 3 years ago

d9d7d5e2

[Static Runtime] Fix bug in quantized::linear wrapper (#59407)

Hao Lu committed 3 years ago

6627c00e

[NNC] Fix BufHandle arguments in loopnest python API (#59348)

huiguoo committed 3 years ago

7d38901e

[torch distributed] Implementing reduce_scatter_base (#57567)

liangluofb committed 3 years ago

77de640f

Revert D28859795: [nnc] Enable CPU fusion inside Facebook, take 4

bertmaher committed 3 years ago

46d724c9

Update reviewer list for the distributed package (#59417)

cbalioglu committed 3 years ago

526445df

Prefer accurate reciprocal on ARMv8 (#59361)

malfet committed 3 years ago

aa4f27c1

Automated submodule update: FBGEMM (#59337)

facebook-github-bot committed 3 years ago

3416b8dd

Fix the "tensors to be on the same device" error in HistogramObserver (#59234)

KAI ZHAO committed 3 years ago

1aa14fcb

Support switching RemoteModule between train/eval (#59026)

Lishan Yang committed 3 years ago

2aa463d9

Revert D28538996: Enable torch::deploy GPU tests in sandcastle

H-Huang committed 3 years ago

c1c9774a

Add build support for kineto + rocm (#58401)

mwootton committed 3 years ago

e66015da

Newer Older