strisunshinewentingwang/pytorch

Pull Requests Commits

ci: Remove CUDA 11.1 binary builds

seemethere committed 2 years ago

67cb0f2a

Manual skip sparse tests

rohan-varma committed 2 years ago

c3d79ac4

[FSDP][BE] remove get_full_params() from test code (#73242)

rohan-varma committed 2 years ago

199d1cb9

[FSDP] summon_full_params fix (#73314)

rohan-varma committed 2 years ago

e10cd886

[jit][edge] Fix array index checking in mobile interpreter. (#73241)

zhxchen17 committed 2 years ago

cafd0f33

[PyTorchEdge] Extend flatbuffer to support extra files map (#72951)

pavithranrao committed 2 years ago

62eb7d64

[JIT] add builtin functions for (complex, Tensor) (#73286)

davidberard98 committed 2 years ago

f51a497f

[Pytorch Edge] Lean Runtime Test

JacobSzwejbka committed 2 years ago

faacb8ab

set SCCACHE_IGNORE_SERVER_IO_ERROR=1

mszhanyi committed 2 years ago

3f7c17a2

Try retryable steps for installing nvidia tools

janeyx99 committed 2 years ago

9b271604

Fix serialization and deepcopying for wrapper subclasses

jbschlosser committed 2 years ago

30653d16

[JIT] Add GRAPH_DEBUG for conv_add_relu fusion (#73138)

davidberard98 committed 2 years ago

8af23ea0

Revert D33994011: Make debug_pkl smaller by only emitting unique traces.

albanD committed 2 years ago

3bd1507f

[Static Runtime] Avoid checks during op execution for TupleConstruct & ListConstruct (#69029)

d1jang committed 2 years ago

5772b1af

Revert D34400588: [pytorch][PR] super setUp call missing in TestSparse

albanD committed 2 years ago

49444bb5

Introduce an environment variable to change c10 log level (#71746)

cbalioglu committed 2 years ago

7366724e

Make debug_pkl smaller by only emitting unique traces. (#72596)

qihqi committed 2 years ago

3d37f5b0

Check clang++/g++ version when compiling CUDA extensions (#63230)

andfoy committed 2 years ago

86deecd7

Update ONNX submodule to 1.11.0 (#73111)

garymm committed 2 years ago

3ac78281

Move Lazy Shape Inference functions to pytorch core (#72756)

wconstab committed 2 years ago

d56d530e

[torch] do not fold bmm into mm when tensor1 dim==3 but not contiguous (#73115)

jspark1105 committed 2 years ago

fffb97f3

[PyTorch] [Model Tracer] Protect writes to set/map for capturing used operators, dtypes, custom classes, and build features (#73327)

dhruvbird committed 2 years ago

f4611ee9

[PyTorch] [Model Tracer] Refactor method bodies in various record function handlers into source (.cpp files) (#73321)

dhruvbird committed 2 years ago

c2279d48

Fix tensor.__deepcopy__ for lazy device #73197 (#73216)

wconstab committed 2 years ago

Verified f2da2e15

Enable CSR inputs for torch.sparse.mm (#73075)

IvanYashchuk committed 2 years ago

ebd93f69

Add dynamic shape support to AOT driver & compiler (#72995)

priyaramani committed 2 years ago

ac97e953

Squeeze p2: hook up Squeeze to LazyView (#73067)

Krovatkin committed 2 years ago

5a7778c9

[shard] fix some imports in tests (#73309)

wanchaol committed 2 years ago

6c6ae0e7

[shard] Extensible ShardingSpec (#72130)

wanchaol committed 2 years ago

d6c5295e

[vulkan] Remove dead code from previous Vulkan backend (#73243)

SS-JIA committed 2 years ago

b3e3eb99

Newer Older