strisunshinewentingwang/pytorch

Pull Requests Commits

[profiler] record nccl version in distributed info (#121044)

shengbao-zheng committed 250 days ago

eea37c6d

[Clang-tidy header][23/N] Enable clang-tidy coverage on aten/src/ATen/*.{cpp,h} (#121380)

cyyever committed 250 days ago

3aa512cd

[dynamo] relax missing symbols runtime assert (#121339)

IvanKobzarev committed 250 days ago

9a450019

[Inductor] Allocate another shard for testing cpp-wrapper JIT (#121310)

desertfire committed 250 days ago

0339f1ca

[Inductor] Enable ABI-compatible mode for cpp-wrapper JIT (#121309)

desertfire committed 250 days ago

7e598c00

[ Inductor ] Shape padding honors output stride preservation (#120797)

kadeng committed 250 days ago

57fc35a3

Change ATEN generator argument type to const std::optional<Generator>& (#120076)

cyyever committed 250 days ago

4305c64f

[inuctor] fix the layout problem for nll_loss2d_backward (#121173)

shunting314 committed 250 days ago

1ce50496

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

mingfeima committed 250 days ago

b3065f68

[FSDP2] Relaxed check for parent mesh (#121360)

awgu committed 250 days ago

e8e3049f

Add SDPA pattern for HuggingFace models BF16 (#121202)

Valentine233 committed 250 days ago

db36d21f

Wrap remote cache creation with a try-catch (#121340)

oulgen committed 250 days ago

953c6c37

Modify StorageImplCreateHelper (#118459)

CLiqing committed 251 days ago

291ce86a

[Quant][Inductor] Fix q/dq per channel lowering with 64-bit qparams (#120984)

Xia-Weiwen committed 251 days ago

f848e9c6

[DTensor][XLA] refactor DTensor _xla API (#113214)

yeounoh committed 251 days ago

4f9d4e1a

[CUDACachingAllocator] Simplify update_stat and avoid casts (#120964)

cyyever committed 251 days ago

c723514e

Make CausalBias a torch.Tensor subclass again (#121358)

drisspg committed 251 days ago

55232c4e

[dtensor][debug] have visualize_sharding correctly print for sub-mesh DTensor (#121216)

XilunWu committed 251 days ago

df2ad1fe

[dtensor][1/N] add torchrec even row-wise sharding example (#120260)

XilunWu committed 251 days ago

77873f6f

[dtensor][debug] allow visualize_sharding to print header (#121179)

XilunWu committed 251 days ago

9cc0f23e

Bugfix consume_prefix_in_state_dict_if_present function to keep the order of the state_dict (#117464)

jmarintur committed 251 days ago

a2854ae9

Prevent infinite recursion within Tensor.__repr__ (#120206)

aorenste committed 251 days ago

edd80f87

graph break on sparse tensors constructions (#120458)

laithsakka committed 251 days ago

eb4d87f2

[TP] Introduce Sequence Parallel Style for Laynorm/RMSNorm/Dropout (#121295)

wanchaol committed 251 days ago

1a28ebff

[cuDNN] Cleanup cuDNN < 8.1 ifdefs (#120862)

eqy committed 251 days ago

967dd316

[profiler] Add execution_trace_observer as an optional argument to profiler (#119912)

briancoutinho committed 251 days ago

b9087f85

[DCP] Adds main in format utils (#120128)

LucasLLC committed 251 days ago

eb114543

Use c10::irange and fix other index types in ForeachReduceOp.cu (#121123)

cyyever committed 251 days ago

5cc511f7

Make ATen-cpu cuda/rocm agnostic (#121082)

xw285cornell committed 251 days ago

c268ce4a

Use type check for also `is_not` (#113859)

oraluben committed 251 days ago

e50ded03

Newer Older