strisunshinewentingwang/pytorch

Pull Requests Commits

Add sparse compressed fake tensor support (#120920)

pearu committed 200 days ago

ce290308

Add a decomposition for torch.put, 2. (#120179)

pearu committed 200 days ago

c0649998

Fix compilation error: load_fp32_from_fp16’ was not declared in this scope for ppc64le (#120307)

pratiklp00 committed 200 days ago

8ba49d0e

Fix hipification issue (#121107)

jianyuh committed 200 days ago

27ac7307

[dtensor] change distribute_module input/output_fn to accept module (#120895)

wanchaol committed 200 days ago

2e505667

Do not use warm_pool() if TorchTnT is used (#121047)

ckluk2 committed 200 days ago

3045b164

Fix comparison of integer expressions of different signedness (#121066)

cyyever committed 201 days ago

4b494d07

[PT2][Inductor] Fix missing "example_value" for nodes introduced by group batch fusion (#120974)

mengluy0125 committed 201 days ago

c83dfc88

[jit][nested strided tensor] support nested tensor in check_trace (#121039)

davidberard98 committed 201 days ago

cead0363

If data dependent, check if guard_size_oblivious would fix problem and report if so (#121011)

ezyang committed 201 days ago

089f4c0b

[Clang-tidy header][21/N] Fix clang-tidy warnings in aten/src/ATEN/*.{cpp,h} (#120763)

cyyever committed 201 days ago

13fadea8

[inductor] add decompostition for mm in backward (#120933)

jackiexu1992 committed 201 days ago

4f0481e1

[dynamo][compile-time] Remove unnecessary tree_map_only (#121052)

anijain2305 committed 201 days ago

b7f25226

Revert "[PT2D] Make the speedup benchmark works with DDP + CompiledAutograd (#120454)"

pytorchmergebot committed 202 days ago

368f242e

[dynamo] Minor refactors (#120966)

jansel committed 202 days ago

0e0a621e

[dynamo][comp-time] BuiltinVariableTracker - inspect signature only on failure (#121053)

anijain2305 committed 202 days ago

8e430107

[DCP] deletes legacy formatting test (#120127)

LucasLLC committed 202 days ago

7aced61c

[dynamo][guards-cpp-refactor] Skip type and length check guard for DictGuardManager (#120739)

anijain2305 committed 202 days ago

7f81563e

[dynamo][guards-cpp-refactor] DICT_CONTAINS guard (#120673)

anijain2305 committed 202 days ago

82d1465d

[dist][sharded_tensor] Fix ChunkShardingSpec metadata offsets for empty shards (#121002)

IvanKobzarev committed 202 days ago

bab4b5a3

[export][ez] minor variable rename (#121040)

suo committed 202 days ago

66b20b42

[export] cleanup to rewrite steps (#121037)

suo committed 202 days ago

50563719

[Torchelastic][Logging] Pluggable logsspecs using python entrypoints and option to specify one by name. (#120942)

kurman committed 202 days ago

b0cfa96e

remove constraints from capture_pre_autograd_graph (#120981)

avikchaudhuri committed 202 days ago

f351a71d

[Quant][Inductor] Enable lowering of dynamic qlinear for X86Inductor (#120605)

Xia-Weiwen committed 203 days ago

83d848e1

[dtensor] add support for loss parallel (#119877)

tianyu-l committed 203 days ago

af5376c4

[inductor] fix accuracy failure for a few models under freezing (#121054)

shunting314 committed 203 days ago

c4ed456f

add int8 packed gemm support on CPU device (#118056)

mingfeima committed 203 days ago

f84375ca

[ATen-vulkan][EZ] Bug fixes: only create the image view when memory has been bound, invalidate cmd on flush (#121027)

SS-JIA committed 203 days ago

5258c364

Add the bound check for flatten with out_dim (#120894)

lancerts committed 203 days ago

2d9efad3

Newer Older