strisunshinewentingwang/pytorch

Pull Requests Commits

[quant][fx][devs] Add lowering support for torch.cat (#72487)

jerryzh168 committed 2 years ago

ac0cac77

Move fx2trt out of PyTorch core (#72499)

Yinghai Lu committed 2 years ago

36704662

[jiterator] Make USE_JITERATOR error if jit_macros.h isn't included (#71959)

peterbell10 committed 2 years ago

60aea2e4

Automated submodule update: FBGEMM (#72483)

facebook-github-bot committed 2 years ago

1af82001

[quant][fx] Add lowering support for ops in GeneralTensorShapeOpQuantizeHandler (#72387)

jerryzh168 committed 2 years ago

4b69a237

[package] fix orderedimporter dummy package check (#72533)

suo committed 2 years ago

6a1147d0

Beta value is ignored for sparse torch.addmm with non-MKL build (#72430)

IvanYashchuk committed 2 years ago

ad5a5a97

[optim] ASGD fold state updates into functional and pass list of vars rather than states (#71335)

mikaylagawarecki committed 2 years ago

8bb1d067

[optim] NAdam fold state updates into functional (#71334)

mikaylagawarecki committed 2 years ago

ccc1a01d

[SR] Make fused_gather_ranges_to_dense work on graph outputs (#71498)

Mike Iovine committed 2 years ago

41aba71c

[SR] Add native implementations for converted prim ops (#71474)

Mike Iovine committed 2 years ago

6c0521b9

Correcting a minor typo: "Users should pay" instead of "Users should be pay" (#72500)

lefoun committed 2 years ago

8d525d47

Structured Kernels for `index_copy`, add `out` variant (#67329)

krshrimali committed 2 years ago

bc03c1d0

[JIT][easy] Fix off-by-one error in tupleIndex (#72447)

Mike Iovine committed 2 years ago

7d917a52

[ROCM] Navi21 Enablement 1 (#69942)

micmelesse committed 2 years ago

a88c5e6a

[DataPipe] Fixing MapDataPipe docstrings (#72476)

NivekT committed 2 years ago

8886ed2d

[DataPipe] Fixing IterDataPipe docstrings (#72475)

NivekT committed 2 years ago

b4e5b4d9

DBR Quantization: Add support for functional conv variants (#71795)

andrewor14 committed 2 years ago

9d08318a

add fx2trt diagnostics (and a framework) (#72374)

kflu committed 2 years ago

fed0ec30

Show friendly error message when forgetting `init` in `torch.cuda` (#72404)

xuzijian629 committed 2 years ago

c4af6ba1

[Caffe2] Use more irange()s in loops. (#72273)

sweettea committed 2 years ago

e578a801

[Caffe2] Use more irange()s in loops. (#72262)

sweettea committed 2 years ago

fd3ed2e4

[Caffe2] Use more irange()s in loops. (#72265)

sweettea committed 2 years ago

2ea50319

[Caffe2] Use more irange()s in loops. (#72264)

sweettea committed 2 years ago

27f3ce85

[Caffe2] Use more irange()s in loops. (#72270)

sweettea committed 2 years ago

4cca3935

[JIT] enable profiling optional tensors (#70532)

davidberard98 committed 2 years ago

c3147504

[optim] update step in functional and pass state_steps instead of state (#71333)

mikaylagawarecki committed 2 years ago

7176c926

Structured Kernel Precompute codegen handle fields without replacement (#71368)

francescocastelli committed 2 years ago

5e6f2966

#71946 Remove Python 3.6 references (#72211)

musebc committed 2 years ago

8bf3179f

[fx2trt] remove split.py (#71933)

842974287 committed 2 years ago

2afed243

Newer Older