strisunshinewentingwang/pytorch

Pull Requests Commits

fix native_layer_norm meta kernel parity w cuda (#83457)

bdhirsh committed 3 years ago

1ee9eb52

fix resnet50_quantized_qat and mobilenet_v2_quantized_qat <> functionalization (#83339)

bdhirsh committed 3 years ago

f4b7c10e

Revert "Use opmath_type for CUDA logcumsumexp (#83425)"

pytorchmergebot committed 3 years ago

785f7f62

[quant] Remove unused quantize handler definitions (#83360)

jerryzh168 committed 3 years ago

3586af8a

[vulkan] Use aliases when retrieving from packed/unpacked lists in OpContexts (#83526)

ssjia committed 3 years ago

05932146

Add option to run anomaly mode without nan checking (#83481)

soulitzer committed 3 years ago

31fad392

ci: Add workflow to build official docker images with multiarch (#83437)

seemethere committed 3 years ago

1b437718

[Expanded Weights] add 'same' and 'valid' padding support (#83345)

samdow committed 3 years ago

3a511e83

[ONNX] Update the script for version updates (#83283)

justinchuby committed 3 years ago

cd68f089

[ROCm] MIOpen fused convolution relu (#82002)

jeffdaily committed 3 years ago

d52d2bd5

update merge failed msg (#83462)

kshitij12345 committed 3 years ago

79356311

[Nested Tensor] Created Nested Tensor to Nested Tensor Views (#82658)

drisspg committed 3 years ago

4b597019

[maskedtensor] first commit, core and creation (#82836)

george-qi committed 3 years ago

94ba085c

Vectorize cpu tensor conversions (#80905)

peterbell10 committed 3 years ago

84146f3d

Fix _refs.lcm using floating point maths (#82950)

peterbell10 committed 3 years ago

559c8b89

[ROCM] Enable test_memory_format_nn_BatchNorm tests on ROCM (#82512)

micmelesse committed 3 years ago

9745edf9

Use opmath_type for CUDA logcumsumexp (#83425)

peterbell10 committed 3 years ago

06a64f7e

Split ScanKernels.cu (#83422)

peterbell10 committed 3 years ago

0faf10b0

[ROCm] Fixes the kernel asserts API declaration mismatch error (#81790)

pruthvistony committed 3 years ago

8473e696

[primTorch] Add ref for movedim (#83278)

Nikita Karetnikov committed 3 years ago

b156f332

module names are made more consistent with POI page (#83219)

b0noI committed 3 years ago

2c79b9c6

[easy] Fix .sizes() call in saved_variable.cpp for nested tensor (#83356)

mikaylagawarecki committed 3 years ago

92a00588

[functorch] classify some more test failures (#83520)

zou3519 committed 3 years ago

7e7afcab

[functorch] audit skips and xfails for vjp tests (#83518)

zou3519 committed 3 years ago

52b8a581

[functorch] Classify some vmap failures with comments (#83517)

zou3519 committed 3 years ago

64a3fbae

[primTorch] Add ref for diagonal and more test inputs (#82321)

Nikita Karetnikov committed 3 years ago

a3e3cbfb

[primTorch] Fix off by 1 in `canonicalize_dim` (#83198)

Nikita Karetnikov committed 3 years ago

4010f961

Revert "reverted diff: Add python stack tracing option on on-demand flow" (#82378)

slgong-fb committed 3 years ago

6a5ca409

[vulkan][fix] Fix unsafe direct array access (#83432)

ssjia committed 3 years ago

bb94a13d

[vulkan] Replace *_size() functions with get_dim<N>() (#83423)

ssjia committed 3 years ago

08d38bbc

Newer Older