strisunshinewentingwang/pytorch

Pull Requests Commits

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

mengluy0125 committed 200 days ago

c83dfc88

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

davidberard98 committed 200 days ago

cead0363

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

ezyang committed 200 days ago

089f4c0b

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

cyyever committed 200 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 201 days ago

368f242e

[dynamo] Minor refactors (#120966)

jansel committed 201 days ago

0e0a621e

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

anijain2305 committed 201 days ago

8e430107

[DCP] deletes legacy formatting test (#120127)

LucasLLC committed 201 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 202 days ago

83d848e1

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

tianyu-l committed 202 days ago

af5376c4

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

shunting314 committed 202 days ago

c4ed456f

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

mingfeima committed 202 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 202 days ago

5258c364

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

lancerts committed 202 days ago

2d9efad3

[dynamo bug burndown] update tensor creation to support sequences of tensors (#120872)

ColinPeppler committed 202 days ago

06fe6ed8

[Dynamo] Fix guards for code objects (#120909)

malfet committed 202 days ago

a3b81666

Change TestOpWaitiness to use MultiProcessTestCase (#121046)

yifuwang committed 202 days ago

f7a2bae0

[FSDP2] Used `ReduceOp.AVG` if fp32 reduce-scatter (#120919)

awgu committed 202 days ago

4cf6d117

Fix more xfails for scaled_dot_product_attention (#121032)

davidberard98 committed 202 days ago

85157af7

[DTensor] Supported `foreach=True` for `clip_grad_norm_` (#120910)

awgu committed 202 days ago

7c71d7f3

[DTensor] Supported `foreach=False` for `clip_grad_norm_` (#120238)

awgu committed 202 days ago

f0e8e7cf

Newer Older