strisunshinewentingwang/pytorch

Pull Requests Commits

Add support to call unpack for pytorch mobile quantized FC and Conv (#26211)

supriyar committed 5 years ago

d46b982d

flat hash map that preserves insertion and deletion order (#25675)

Elias Ellison committed 5 years ago

921079c5

make copy (#26371)

Elias Ellison committed 5 years ago

43b30cd5

Add per channel observer (#25887)

Haixin Liu committed 5 years ago

dcbfc3bd

Revert D17374409: [pytorch][PR] Implement more size-oriented opcodes in the depickler.

houseroad committed 5 years ago

7042bfea

Export gelu (#24475)

neginraoof committed 5 years ago

293d73fc

Implement more size-oriented opcodes in the depickler. (#25786)

Owen Anderson committed 5 years ago

51275991

Export clamp for opset 11 (#25797)

BowenBao committed 5 years ago

595c1dfa

Revert D17464904: Add NoQEngine to QEngine and refactor the name of set/get qengine

Ailing Zhang committed 5 years ago

b1ecf4bc

Fix quantized::linear QuantFusion patterns (#26414)

jerryzh168 committed 5 years ago

cbc7172a

Add NoQEngine to QEngine and refactor the name of set/get qengine (#26330)

jerryzh168 committed 5 years ago

4f7292f7

add script to build mobile library with host toolchain (#26440)

ljk53 committed 5 years ago

c3f881cd

Back out "[pytorch][PR] Fix many type mismatches in the CUDA version of calc_digamma and calc_trigamma" (#26444)

Mike Ruberry committed 5 years ago

495dbacf

Remove quantizeBias (#26388)

jerryzh168 committed 5 years ago

fb28014a

Exclude libfbjni.so from pytorch_android not to have its duplicating (#26382)

IvanKobzarev committed 5 years ago

6387ffab

tvm operator dynolog (#26295)

ajyu committed 5 years ago

e44ea6cd

Move the CUDA implementation of rsqrt to ATen. (#25285)

VitalyFedyunin committed 5 years ago

36ade9aa

C++ API parity: at::Tensor::is_leaf

pbelevich committed 5 years ago

44ffbc43

fix composite learning rate (#26227)

olittle committed 5 years ago

a8386d2a

Add extra filtering for scale/zero_point/dtype in FoldQuantizeCallIntoBuffer (#26224)

jerryzh168 committed 5 years ago

f75c1e49

Adding quantized::conv2d function for pytorch mobile in c10 (#26152)

supriyar committed 5 years ago

b23be955

remove extra get_worker_id call in distributed rpc init (#26381)

rohan-varma committed 5 years ago

1f510512

Add filter function to subgraph rewriter runGraph (#26223)

jerryzh168 committed 5 years ago

f29e0d70

Use static type information to restore type tags (#25447)

zdevito committed 5 years ago

12762cd5

Add ivalue::type(), part 1 (#25439)

zdevito committed 5 years ago

ad0af112

add pass for onnx scalar type conversion (#24378)

BowenBao committed 5 years ago

d02369da

Adds dtypes decorators to and allows helper methods in device generic test classes (#26375)

Mike Ruberry committed 5 years ago

248d5857

Disable QNNPACK tests if pytorch is not built with it. (#26427)

supriyar committed 5 years ago

52d999e1

Puts ROCm tests on default stream (#26394)

Mike Ruberry committed 5 years ago

a5616602

Fix many type mismatches in the CUDA version of calc_digamma and calc_trigamma (#25791)

xuhdev committed 5 years ago

13b54460

Newer Older