strisunshinewentingwang/pytorch

Pull Requests Commits

[LTC] Update merge_rules.yaml

alanwaketan committed 1 year ago

4d2893cb

Make GenLazyNativeFuncDefinition generator to be customizable in lazy codegen (#87823)

wonjoolee95 committed 1 year ago

0efd4e92

Update all ONNX symbolics with new JitScalarType API (#87245)

Thiago Crepaldi committed 1 year ago

a8f40b39

[vision hash update] update the pinned vision hash (#88382)

pytorchmergebot committed 1 year ago

b013825c

Update pybind11 to v2.10.1 (#88332)

Skylion007 committed 1 year ago

5fb9c113

Improve perf by avoiding implicit string creation in c10_cuda_check_implementation (#88350)

r-barnes committed 1 year ago

e59d307e

[codegen] using TORCH_LIBRARY_FRAGMENT for some namespaces (#88229)

jerryzh168 committed 1 year ago

a0fb234b

Not run inductor test in trunk (#88374)

huydhn committed 1 year ago

7b8cc063

Added add/mul for nested dense [B, *, D], [B, 1, D] case (CUDA-only) (#88289)

mikaylagawarecki committed 1 year ago

d979caa8

Split out forward AD tests from test_ops_gradients and reenable slow gradcheck CI (#88216)

soulitzer committed 1 year ago

4c20c050

Revert "[inductor] Handle the case where kwargs contains tensor (#88215)"

pytorchmergebot committed 1 year ago

a8561c45

[LTC] Remove non-native view ops (#88031)

alanwaketan committed 1 year ago

7354368f

[Pytorch][Vulkan] Update spv generation script to embed shader parameters (#88321)

kimishpatel committed 1 year ago

72f36880

[FSDP][Easy] Remove unneeded `TrainingState` transition (#88232)

awgu committed 1 year ago

6c858e37

[FSDP] Rename `unflat_param_name` -> `fqn` for consistency (#88123)

awgu committed 1 year ago

73de44fc

[FSDP] Simplify `_get_buffer_names()` (#88122)

awgu committed 1 year ago

f35d5145

[FSDP] Remove unneeded `torch.no_grad()` context when offloading to CPU (#88121)

awgu committed 1 year ago

572a3d2d

[FSDP][Docs] Add note mentioning rate limiter for backward prefetch (#88120)

awgu committed 1 year ago

c87f0501

[FSDP()][27/N] Add forward hook registration (#88040)

awgu committed 1 year ago

32d22edc

Add _foreach_addc(div/mul)(_).Tensor (#88157)

cpuhrsch committed 1 year ago

6fd41665

[ONNX] Produce comprehensive assertion errors for quantized outputs (#87242)

jazzysoggy committed 1 year ago

91a51fe9

[1/n] Thread PG: fix pyre error of class ProcessGroup (#88281)

yhcharles committed 1 year ago

ca2dc8b4

Update Reviewers for CPU-related Modules (#87591)

jgong5 committed 1 year ago

d1ba4c3a

[nvFuser] patches profiling on scalar arguments for std/var (#88165)

jjsjann123 committed 1 year ago

b325c3fc

Revert "torchdynamo support modules() for nn_module (#88023)"

pytorchmergebot committed 1 year ago

bf7c996d

Print only the driver version from the first GPU (#88364)

huydhn committed 1 year ago

7dfa7554

Use tensor cores for NT bmm (#86856)

cpuhrsch committed 1 year ago

943b20e7

[PyTorch] Make c10::irange(x) generate the same assembly as for loop (#86841)

swolchok committed 1 year ago

1c0d47cb

Add [[noreturn]] attribute to operator() in DispatchKeyExtractor.h (#88333)

zou3519 committed 1 year ago

ef4ce6d4

[inductor] Handle the case where kwargs contains tensor (#88215)

desertfire committed 1 year ago

983c0e7f

Older