strisunshinewentingwang/pytorch

Pull Requests Commits

[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

Add batching rule for "is_complex", "conj" (#44649)

zou3519 committed 4 years ago

1cd5ba49

Add bound method tests for async_execution with RRef helper (#44716)

mrshenli committed 4 years ago

cce7680a

Make async_execution compatible with RRef helpers (#44666)

mrshenli committed 4 years ago

257c6d0f

Add _get_type() API to RRef (#44663)

mrshenli committed 4 years ago

924717bf

Vec256 Test cases (#42685)

quickwritereader committed 4 years ago

6954ae12

fixes lda condition for blas functions, fixes bug with beta=0 in addmv slow path (#44681)

ngimel committed 4 years ago

e6101f55

Remove many unused THC pointwise math operators (#44230)

xuhdev committed 4 years ago

570102ce

Remove EXPERIMENTAL_ENUM_SUPPORT feature guard (#44243)

gmagogsfm committed 4 years ago

07d07e3c

Reference amp tutorial (recipe) from core amp docs (#44725)

mcarilli committed 4 years ago

3e6bb523

change self.generator to generator (#44461)

thuzhf committed 4 years ago

a011b861

Newer Older