strisunshinewentingwang/pytorch

Pull Requests Commits

[ROCm] Enable Bfloat16 type for activation and batch-norm

rohithkrn committed 4 years ago

66ee4f1c

Drop support of the build option USE_GLOO_IBVERBS (#33163)

xuhdev committed 4 years ago

f255b7a3

add missing default value for LRScheduler.step() (#32411)

christopher-hesse committed 4 years ago

1487137c

Fix link to py-spy content in contribution guide TOC (#31760)

ngoldbaum committed 4 years ago

139afd0e

Revert D19825127: [pytorch][PR] Move where cuda implementation to TensorIterator

Saurabh Aggarwal committed 4 years ago

74c8a8f7

bad tbb lambda capture, bad chunk size (#30352)

EscapeZero committed 4 years ago

000a5e2b

Quantized leaky relu

z-a-f committed 4 years ago

a23009f9

Build ahead-of-time C++ extensions with ninja on windows

peterjc123 committed 4 years ago

769abddf

TorchScript add check if quantized

z-a-f committed 4 years ago

acd51e13

Use C10_WARP_SIZE to fix functionality on HIP vs CUDA for batch_norm_backward_reduce (#33098)

jithunnair-amd committed 4 years ago

cb39a540

[ONNX] Fix ONNX CI (#33200)

houseroad committed 4 years ago

44723a1c

Temporarily disable failing 'binary_macos_libtorch_2_7_cpu_build' and… (#33207)

kostmo committed 4 years ago

af4d6120

Update CPU threading doc (#33083)

Ilia Cherniavskii committed 4 years ago

04829e92

Prepare templates (#30982)

izdeby committed 4 years ago

6706c3f4

Remove some Half support in some binary CPU kernels (#33021)

xuhdev committed 4 years ago

45818a3d

optimize cat performance on CPU with TensorIterator (#30806)

mingfeima committed 4 years ago

7b50e762

Removes flaky check (#33146)

Mike Ruberry committed 4 years ago

ad90c97c

Use int64 in pdist kernel to handle batches >= 46342 #30583 (#31593)

ptrblck committed 4 years ago

a64d0ffe

Move where cuda implementation to TensorIterator (#32984)

zasdfgbnm committed 4 years ago

367488b0

Add zero_mask function for vectorized functions. (#32985)

xuhdev committed 4 years ago

31370949

Revert D18749922: [pytorch] Migrating index_add cuda to ATen

zhangguanheng66 committed 4 years ago

855ee644

Updated DispatchKeyExtractor to expect TensorOptions (#30981)

izdeby committed 4 years ago

857bae39

Remove return value for __exit__ (#32997)

vincentqb committed 4 years ago

e7f0b154

[caffe2] use JIT'ed fp32 SLS (#33123)

jspark1105 committed 4 years ago

6c0dc66c

Add allow_rebase_history flag and fix codegen functions for multiple views (#32790)

albanD committed 4 years ago

36559755

[pytorch] Migrating index_add cuda to ATen (#30573)

ggoossen committed 4 years ago

330d051b

fix gather regression by not materializing loop vars in the error mes… (#33108)

ngimel committed 4 years ago

9857d9b4

[1/3] Bind IndexHash to PyTorch (#33015)

idning committed 4 years ago

6f46962f

Updating submodules

svcscm committed 4 years ago

61ac14a4

Use bazelisk instead of specifying bazel version manually. (#33036)

Ailing Zhang committed 4 years ago

a3e69d34

Newer Older