strisunshinewentingwang/pytorch

Pull Requests Commits

Revert D15248618: Split ATen/Parallel into interface and backend

ezyang committed 5 years ago

fd95947e

Some minor fix to unblock the Bert model quantization (#20787)

lly-zero-one committed 5 years ago

70ecddfd

Add quant-dequant nodes for bias. (#20045)

nishantpdce committed 5 years ago

a501e7d5

Add DictType to Metadata (#20770)

houseroad committed 5 years ago

c2d0e731

Use AT_INTERNAL_ASSERT in test_base (#20555)

suo committed 5 years ago

70eb315d

Rename FC to Linear in the test routine (#20716)

jianyuh committed 5 years ago

4a85e795

fbgemm precision argument (#20790)

jerryzh168 committed 5 years ago

77651615

Split ATen/Parallel into interface and backend (#20057)

Ilia Cherniavskii committed 5 years ago

c4a3b4d5

int_repr for different quantized types (#20656)

jerryzh168 committed 5 years ago

adbab828

Use SmallVector to allocate Compound operands inline. (#20762)

Owen Anderson committed 5 years ago

c1d6bcf3

Optimize pytorch layer_norm forward (#20345)

xiaomengy committed 5 years ago

c9da0119

use tensoriterator instead of th for fill_ implementation. (#20719)

nairbv committed 5 years ago

9cec8ae1

Fix copy_transpose_valid check (#20759)

colesbury committed 5 years ago

7a0c6d52

make magic methods work with casts too (#20654)

Elias Ellison committed 5 years ago

5acc664f

Change Bias to QTensor with qint32(int32_t) (#20713)

jianyuh committed 5 years ago

e6f22e1b

Change Weight to QTensor with qint8(int8_t) (#20712)

jianyuh committed 5 years ago

b9a150ed

Fix a bug in quantize_linear (#20711)

jianyuh committed 5 years ago

ac2314fd

Update Conda description in PyTorch README (#20726)

dreiss committed 5 years ago

32803b52

Auto-convert GPU arrays that support the __cuda_array_interface__ protocol (#20584)

madsbk committed 5 years ago

5d8879cf

Improve the recommended citation

xuhdev committed 5 years ago

847d9c57

Add support for CMake switches for VS 2019 (#20752)

peterjc123 committed 5 years ago

bb20956e

add str comparisons (#20761)

Elias Ellison committed 5 years ago

47dc65fe

Add dequantize_linear for JIT pass (#20107)

jerryzh168 committed 5 years ago

cca923c4

Throw error if multiple kernels registered (#20737)

smessmer committed 5 years ago

cc02a1af

Hipify fb/quantize

xw285cornell committed 5 years ago

f3d827f3

Split cpu/gpu in caffe2/distributed + some clean up (#20674)

xw285cornell committed 5 years ago

b5edeca3

compile with -fcolor-diagnostics (#20662)

suo committed 5 years ago

d7cd2d7a

Fix missing cudatoolkit dependency in binary linux tests

pjh5 committed 5 years ago

c790f10e

Fixing upload_binary_htmls again

pjh5 committed 5 years ago

e3970d66

Revert D15178352: [pt1][quant] Quantized Conv2d operator

pjh5 committed 5 years ago

fac307a5

Newer Older