strisunshinewentingwang/pytorch

Pull Requests Commits

Move mobile specific CPUCachingAllocator to c10/mobile folder. (#45364)

kimishpatel committed 4 years ago

6e55a26e

Updates deterministic flag to throw a warning, makes docs consistent (#45410)

Mike Ruberry committed 4 years ago

b2925671

Conservative-ish persistent RNN heuristics for compute capability 8.0+ (#43165)

mcarilli committed 4 years ago

aa2bd7e1

Updated `cholesky_backward` for complex inputs (#45267)

IvanYashchuk committed 4 years ago

f47fd0eb

Support bfloat16 and complex dtypes for logical_not (#43537)

xuhdev committed 4 years ago

15f85eea

Fix model_name not logged properly issue. (#45488)

xcheng16 committed 4 years ago

ea59251f

[JIT] Enable @unused syntax for ignoring properties (#45261)

Meghan Lele committed 4 years ago

09b3e16b

Add mobile_optimized tag to optimized model. (#45479)

Akshit Khurana committed 4 years ago

5f49d14b

[vulkan][android][test_app] Add test_app variant that runs module on Vulkan (#44897)

IvanKobzarev committed 4 years ago

17be7c6e

[android][vulkan] Module load argument to specify device cpu/vulkan (#44896)

IvanKobzarev committed 4 years ago

2c300fd7

Reorganized Sorting.cpp method order (#45083)

heitorschueroff committed 4 years ago

fe9019cb

Revert D23789657: [wip] fast typeMeta/ScalarType conversion approach 2

Mike Ruberry committed 4 years ago

ab5edf21

Enable torch.cuda.amp typechecking (#45480)

malfet committed 4 years ago

b3135c20

Add cusolver guard for cuda >= 10.1.243 (#45452)

xwang233 committed 4 years ago

df0de780

Improves fft doc consistency and makes deprecation warnings more prominent (#45409)

Mike Ruberry committed 4 years ago

bb19a554

Auto set libuv_ROOT env var for Gloo submodule on Windows platform (#45484)

gunandrose4u committed 4 years ago

0a38aed0

Makes rdiv consistent with div (#45407)

Mike Ruberry committed 4 years ago

6d37126a

Add check for Complex Type to allow non integral alpha. (#45200)

hlahkar committed 4 years ago

7cde662f

Optimize view_as_complex and view_as_real (#44908)

peterbell10 committed 4 years ago

0806c58e

Updates torch.floor_divide documentation to clarify it's actually torch.trunc_divide (or torch.rtz_divide) (#45411)

Mike Ruberry committed 4 years ago

87f98a5b

Missing tests about torch.xxx(out=...) (#44465)

antocuni committed 4 years ago

37f9af7f

Revert D23966878: [pytorch][PR] This PR flips a switch to enable PE + TE

Mike Ruberry committed 4 years ago

56af1226

[wip] fast typeMeta/ScalarType conversion approach 2 (#44965)

Basil Hosmer committed 4 years ago

1ed1a2f5

[quant] Add quant APIs to save/load observer state_dict (#44846)

supriyar committed 4 years ago

489af4dd

[quant] torch.max_pool1d (#45152)

z-a-f committed 4 years ago

bb478810

[TensorExpr] Remove buf_ field from class Tensor. (#45390)

Mikhail Zolotukhin committed 4 years ago

b86008ab

[TensorExpr] Rename `Buffer` to `Placeholder`. (#45389)

Mikhail Zolotukhin committed 4 years ago

3c33695a

[TensorExpr] Consolidate {buffer,function,tensor}.{h.cpp} in tensor.{h,cpp}. (#45388)

Mikhail Zolotukhin committed 4 years ago

92306b85

replaced whitelist with allowlist (#45260)

Vishal487 committed 4 years ago

d2623da5

Add more tests for mt optimizers (#45475)

Iurii Zdebskyi committed 4 years ago

8c309fc0

Newer Older