strisunshinewentingwang/pytorch

Pull Requests Commits

[inductor] Use a large tolerance for botnet26t_256 (#90383)

desertfire committed 3 years ago

f7cdd3a7

[Dynamo] Fix llvm target for meta schedule & add torch to tvm ndarray helper func (#90214)

YJ Shi committed 3 years ago

2b0b4bb6

Fix issue 38095 TODO in test_autograd.py (#90031)

kit1980 committed 3 years ago

6a7659f3

[vmap] Prepend "legacy" to files for old vmap implementation (#90324)

zou3519 committed 3 years ago

4b105349

Make Transformers compilable by C++17 (#90389)

malfet committed 3 years ago

94d800ff

[pruning][core][feature] Implement prune for structured pruning (#89777)

jcaip committed 3 years ago

3531e443

[quant]Fix public bindings for DTypeWithConstraint (#90315)

jcaip committed 3 years ago

d680ea7e

Add hooks structure for passing around user provided hooks, add a new guard_failure_fn (#90371)

voznesenskym committed 3 years ago

4cdc96fb

[BE][CI] Add windows test run instructions (#90388)

malfet committed 3 years ago

c92cf6be

[Quant][fx][bc-breaking] Make convert.py smaller (#90189)

andrewor14 committed 3 years ago

824641b0

reland #89243: [Composable API] replicate: add support for DDP args (#90255)

yhcharles committed 3 years ago

99fb39f5

Give std/var correction overloads proper defaults (#56398)

peterbell10 committed 3 years ago

e6a72787

[MPS] Fix median_out_mps caching (#90326)

malfet committed 3 years ago

b0bd5c45

Reformat optim import (#90294)

fduwjj committed 3 years ago

85ae28b4

[ROCm] Enable few test_prim UTs for ROCm (#88983)

pruthvistony committed 3 years ago

15949fc2

[inductor] More correct check for fbcode environment (#90312)

bertmaher committed 3 years ago

26d1dbc4

Fix exception causes all over the codebase (#90271)

cool-RR committed 3 years ago

351d73b9

[Dynamo+FSDP] Update benchmarks with use_orig_params=True (#90100)

davidberard98 committed 3 years ago

8f079b89

[Dynamo][Easy] capture more exceptions when import skip modules (#90338)

yanboliang committed 3 years ago

898b46d6

[Inductor] More robust stride and offset extraction from index expressions (#90184)

peterbell10 committed 3 years ago

71f27f76

[Inductor] Add test for Scheduler fusions (#90014)

peterbell10 committed 3 years ago

4f448779

[Quant][fx][bc-breaking] Remove unused functions in fx/utils.py (#90025)

andrewor14 committed 3 years ago

13fcc412

Revert "[MPS] Fix median_out_mps caching (#90326)"

pytorchmergebot committed 3 years ago

f28927e9

[quant] Add fused "q - qlinear - dq" operator with skipped quant op for output of linear (#89882)

jerryzh168 committed 3 years ago

887249b2

[Vulkan] Partially fix and then disable copying of vulkan quantized tensors to cpu (#90275)

manuelcandales committed 3 years ago

22e36334

[MPS] Fix median_out_mps caching (#90326)

malfet committed 3 years ago

23c192c3

[fx][passes] Implement annotate getitem node FX passes (#90237)

lequytra committed 3 years ago

b7690059

[quant][fx] Add support for matching constant in the custom matcher code in quantization (#90092)

jerryzh168 committed 3 years ago

0e182c94

as_strided: Fix default storage_offset for reference implementation (#89513)

peterbell10 committed 3 years ago

5caa27a3

Ensure that we fakeify tensor subclasses when they are initially tracked (#90009)

ezyang committed 3 years ago

3d4b92b1

Newer Older