strisunshinewentingwang/pytorch

Pull Requests Commits

[MPS] Fix for matmul errors in test consistency (#83124)

kulinseth committed 2 years ago

02cfefb4

[ONNX] Look at owningBlock instead of graph when recording autograd subgraph (#82852)

BowenBao committed 2 years ago

cb2cb940

Added list clearing codegen to AOTAutograd (hidden behind config.aot_clear_list (#83137)

Chillee committed 2 years ago

ea51e87b

Fix building with Werror (#83275)

peterbell10 committed 2 years ago

d8d9ecbf

Rename PyFunctionPreHook to PyFunctionTensorPreHook (#83225)

soulitzer committed 2 years ago

ccb7d56a

Ensure full fetch before running git commands (#83352)

albanD committed 2 years ago

67e00678

fix a comment since the options in arg parser no longer require Declarations.yaml (#83337)

Krovatkin committed 2 years ago

88d7322b

[ONNX] Add `acceptable_error_percentage` to backend tests (#82622)

BowenBao committed 2 years ago

404c1c04

[ONNX] Move caffe2 import into `exportTest` in `test_models.py` (#82621)

BowenBao committed 2 years ago

7896621f

Switch to the new Windows AMI with conda preinstalled (#83188)

huydhn committed 2 years ago

8598a866

[ONNX] Fix float point detection for optional tensor (with unknown rank) within a list (#81386)

Thiago Crepaldi committed 2 years ago

2c089290

Use merge base to compute number of lines in PR (#83344)

albanD committed 2 years ago

bf21ebce

Fix and unskip dataloader tests on ARM (#83125)

soulitzer committed 2 years ago

4f00c758

[PyTorch] Minor improvement in split_with_sizes (#83208)

swolchok committed 2 years ago

83341329

Support regex-style matching for Any and Oneof (#82853)

SherlockNoMad committed 2 years ago

39e62387

assert that ProxyTensorMode does not accidentally bake in constants (#83297)

bdhirsh committed 2 years ago

0cd8526b

[ONNX] Enable test_uninitialized_optional (#83183)

titaiwangms committed 2 years ago

e327cc5e

Add *_only and all/any pytree utilities (#83316)

ezyang committed 2 years ago

b8b54ecc

Add data_dependent_output tag; generalize proxy tensor to test it (#83312)

ezyang committed 2 years ago

d4237226

Don't build nvfuser benchmarks by default (#67857)

peterbell10 committed 2 years ago

d07a9ba1

Fix internal type conversions in floor_divide and trunc_divide (#83288)

peterbell10 committed 2 years ago

42b572bb

[Quant] Make quantizable LSTM scriptable (#83304)

andrewor14 committed 2 years ago

f81b4ae5

[vision hash update] update the pinned vision hash (#82560)

pytorchmergebot committed 2 years ago

d5fb4e2a

[xla hash update] update the pinned xla hash (#83319)

pytorchmergebot committed 2 years ago

86759ad4

Implement QAT for APoT (#83282)

asl3 committed 2 years ago

0e0f8fd0

An improved version of subgraph matcher (#82090)

SherlockNoMad committed 2 years ago

2ca721cd

Temporarily disable test_fx for dynamo (#83307)

huydhn committed 2 years ago

59b1c4e5

Integrate xdoctest - Rebased (#82797)

Erotemic committed 2 years ago

4618371d

fix functionalization <> resnet18, make ProxyTensor work with tensor-less decomps (#83207)

bdhirsh committed 2 years ago

ba90c9f2

[WIP] Add pr sanity check workflow (#83295)

albanD committed 2 years ago

568c6fb9

Newer Older