strisunshinewentingwang/pytorch

Pull Requests Commits

Update on "refactor intialize bucket views"

zhaojuanmao committed 4 years ago

48126b84

enable type check common_distributed (#44821)

alanashine committed 4 years ago

ba6534ae

Mention TF32 on related docs (#44690)

zasdfgbnm committed 4 years ago

e48201c5

[JIT] Improve Future subtype checking (#44570)

Meghan Lele committed 4 years ago

79108fc1

[FX] Further sanitize generated names (#44808)

James Reed committed 4 years ago

29664e6a

[NNC] Add simplification of Loop + Condition patterns. (#44764)

nickgg committed 4 years ago

204f985f

[pytorch][vulkan][jni] LiteModuleLoader load argument to use vulkan device

IvanKobzarev committed 4 years ago

8ec6bc72

[caffe2] adds Cancel to SafeDequeueBlobsOp and SafeEnqueueBlobsOp (#44495)

dahsh committed 4 years ago

0ccc38b7

[pytorch][bot] update mobile op deps (#44700)

ljk53 committed 4 years ago

3fa7f515

Fix misuse of PyObject_IsSubclass (#44769)

gmagogsfm committed 4 years ago

6befc094

[JIT] Disallow plain Optional type annotation without arg (#44586)

Meghan Lele committed 4 years ago

43fe0345

[NCCL] Add option to run NCCL on high priority cuda stream (#43796)

mingzhe09088 committed 4 years ago

574f9af1

Move `torch/version.py` generation to cmake (#44577)

suo committed 4 years ago

161490d4

[JIT] Disallow plain Tuple type annotation without arg (#44585)

Meghan Lele committed 4 years ago

ffe127e4

enable mypy check for jit_metaprogramming_utils (#44752)

qxu committed 4 years ago

09a84071

[quant] Support clone for per channel affine quantized tensor (#44573)

jerryzh168 committed 4 years ago

3f5bb2ba

[TensorExpr] Support boolean in simplifier (#44659)

asuhan committed 4 years ago

7b3432ca

Vectorize complex copy. (#44722)

ezyang committed 4 years ago

ac0d13cc

[JIT] Disallow plain List type annotation without arg (#44584)

Meghan Lele committed 4 years ago

78b806ab

[JIT] Disallow plain Dict type annotation without arg (#44334)

Meghan Lele committed 4 years ago

cb3b8a33

Add TORCH_SELECTIVE_NAME to AMP definitions (#44711)

ezyang committed 4 years ago

5027c161

[NNC] Fix masking for all block and thread dimensions in CudaCodeGen (#44733)

nickgg committed 4 years ago

82ab167c

[FakeLowP] Addressing FakeLowP OSS issues. (#44819)

venkatacrc committed 4 years ago

a3835179

Fix error code checks for triangular_solve (CPU) (#44720)

IvanYashchuk committed 4 years ago

07d9cc80

Move the description comment of compute_bucket_assignment_by_size from cpp to the header file. (#44703)

Yi Wang committed 4 years ago

f3bd984e

Remove py2 compatible future imports (#44735)

zasdfgbnm committed 4 years ago

20ac7362

[vulkan] glsl shaders relaxed precision mode to cmake option (#43076)

IvanKobzarev committed 4 years ago

6debe825

[FX][EZ] Allow constructing GraphModule with dict for root (#44679)

James Reed committed 4 years ago

e9c6449b

[Caffe2] gcs_cuda_only is trivial if CUDA not available (#44578)

malfet committed 4 years ago

1718b16d

Enable torch.backends.quantized typechecks (#44794)

malfet committed 4 years ago

c44e4878

Older