strisunshinewentingwang/pytorch

Pull Requests Commits

Add Initial NNC Dynamic Shapes Flow (#66136)

Elias Ellison committed 2 years ago

fc82ad18

Adding support for Symbolic Shapes in Inplace Ops #65642 (#65729)

Gamrix committed 2 years ago

26615074

ci: Migrate hardcoded docker builds to GHA (#67455)

seemethere committed 2 years ago

d0bc01fa

Adding `optimizer` attribute to SequentialLR (#67406)

yiwen-song committed 2 years ago

6696c59a

[SR] Native implementation for aten::size (#67346)

Mike Iovine committed 2 years ago

354363b5

[PyTorch][easy] Deduplicate memory planner creation code (#67265)

swolchok committed 2 years ago

9f01937c

Revert D31894777: [pytorch][PR] Replace issue templates with new issue forms

NicolasHug committed 2 years ago

82c35650

[SR] Native implementation for aten::view (#67341)

Mike Iovine committed 2 years ago

afb84344

[jit][edge] Implement torch::jit::Function for mobile funciton. (#65970)

zhxchen17 committed 2 years ago

60472594

[jit] Replace get_executor() with call() in abstract Function interface. (#65969)

zhxchen17 committed 2 years ago

5ef62c88

[SR][C2][easy] Benchmarks report # of ops (#67436)

Mike Iovine committed 2 years ago

8363da3f

[ROCm] Enable frexp support for ROCm builds (#67226)

dllehr-amd committed 2 years ago

b8f07689

[jit] Clean up unneeded virtual methods from Function interface. (#65968)

zhxchen17 committed 2 years ago

07957353

Skip complex128 dtype for test_addmm_sizes_all_sparse_csr Windows test (#67453)

IvanYashchuk committed 2 years ago

bd5e6fe5

Mark mv as CompositeExplicitAutograd (#67373)

Jiewen Tan committed 2 years ago

5b8b2382

Port `tril` and `triu` to structured kernels (#67055)

cbalioglu committed 2 years ago

f3aae629

[qnnpack] Remove asymmetrical padding parameters in qnnpack (#67102)

John Shen committed 2 years ago

4a1f73cc

Formatting changes (#66257)

H-Huang committed 2 years ago

4e873d67

Add FlexiBLAS build support per #64752 (#64815)

Robert Blackwell committed 2 years ago

cee4e8f3

Timing cache for Tensort (#67214)

Shirong Wu committed 2 years ago

55b7387e

Add a Functionalization pass in core (#64432)

bdhirsh committed 2 years ago

0032fa77

add tags for inplace view ops in native_functions.yaml (#65412)

bdhirsh committed 2 years ago

b0a8ca2c

add slice/select/diagonal_scatter variants as primitive ops (#64430)

bdhirsh committed 2 years ago

03f3a033

move some codegen utilities into utils.py (#63094)

bdhirsh committed 2 years ago

665c148e

Back out "Make fb::sigrid_hash_compute_multipler_shift return a std::tuple<int64_t, int64_t>" (#67456)

Mike Iovine committed 2 years ago

b100a9ea

[quant][graphmode][fx][test] Refactor test code for quant-fx2trt unit tests (#67070)

jerryzh168 committed 2 years ago

a8f85300

Add FSDP tests to verify forward overlap and memory usage (#67117)

zhaojuanmao committed 2 years ago

325b1503

Remove process group barrier and all_reduce function calls from tensorpipe agent (#65946)

H-Huang committed 2 years ago

938afa37

Disable linux-xenial-cuda10.2 config (#67344)

malfet committed 2 years ago

0c93c8e3

Document `torch.jit.is_tracing()` (#67326)

kmaehashi committed 2 years ago

6ed68f3f

Newer Older