strisunshinewentingwang/pytorch

Pull Requests Commits

[quant][graphmode][fx] Add support for add relu (#43332)

jerryzh168 committed 4 years ago

26be4dcf

[quant][graphmode][fx] Add support for add (#43331)

jerryzh168 committed 4 years ago

452a4737

.circleci: Prefer using env-file for docker run (#43293)

seemethere committed 4 years ago

6e48c88e

Normalize loops with non-zero start. (#43179)

navahgar committed 4 years ago

100649d6

Revert D23242101: [pytorch][PR] Implement first draft of autograd benchmark.

albanD committed 4 years ago

74781ab5

[quant][graphmode][fx] Add support for conv module + relu (#43287)

jerryzh168 committed 4 years ago

650590da

[quant] Enable from_float for quantized Embedding_Bag (#43176)

supriyar committed 4 years ago

3293fdfa

[quant] Make offsets an optional argument (#43090)

supriyar committed 4 years ago

b354b422

[quant] Create nn.quantized.dynamic.EmbeddingBag (#43088)

supriyar committed 4 years ago

4db8ca11

[TensorExpr] Simplify conditional select (#43350)

asuhan committed 4 years ago

f20a04fa

Fix PackedGemmMatrixFP16 repacking (#43320)

Yinghai Lu committed 4 years ago

743cff4a

Adds arccos, arcsin, arctan aliases (#43319)

Mike Ruberry committed 4 years ago

e57b89c8

Enables bfloat16 x [float16, complex64, complex128] type promotion (#43324)

Mike Ruberry committed 4 years ago

3aec1185

[jit] PyTorchStreamReader::getAllRecord should omit archive name prefix (#43317)

Dmytro Dzhulgakov committed 4 years ago

478fb925

Add Enum convert back to Python object support (#43121)

gmagogsfm committed 4 years ago

0bd35de3

Do not define the macro "isnan" (#43242)

xuhdev committed 4 years ago

f4b6ef9c

Remove erroneous trailing backslashes (#43318)

xuhdev committed 4 years ago

7b520297

Implement first draft of autograd benchmark. (#40586)

albanD committed 4 years ago

c2511bdf

Autograd better error (#43308)

rakshithvasudev committed 4 years ago

0cb52cb4

Add benchmark for performance comparison (#43221)

taivu committed 4 years ago

da036250

[ONNX] Add support for operator `add` between tensor list (#41888)

BowenBao committed 4 years ago

da70976e

Extends test_unary_ufunc.py with numerics, contiguity, domain tests (#42965)

Mike Ruberry committed 4 years ago

c64594f5

Add alias torch.fix for torch.trunc to be compatible with NumPy. (#43326)

hameerabbasi committed 4 years ago

e31cd462

Bias Correction Implementation (#41845)

edmundw314 committed 4 years ago

17f9edda

Replace Conv1d with Conv2d (#42867)

taivu committed 4 years ago

665da61d

Search on system path for Vulkan headers and libraries as a last resort. (#43301)

Ashkan Aliabadi committed 4 years ago

e8139624

[quant] Make OP_LIST_TO_FUSER_METHOD public (#43286)

jerryzh168 committed 4 years ago

217ddea9

Remove proprietary notices

dreiss committed 4 years ago

844d469a

[quant][graphmode][fx] Add support for conv module (#43285)

jerryzh168 committed 4 years ago

9984d335

Reimplement per-operator selective build (#39401)

ezyang committed 4 years ago

7c50c2f7

Newer Older