strisunshinewentingwang/pytorch

Pull Requests Commits

Add Ed to superusers

malfet committed 2 years ago

4267698a

Add Gflags for fusion strategy and make it local to executor (#73668)

Elias Ellison committed 2 years ago

fc362290

[fix] `torch.broadcast_shapes` should not handle shapes with negative dimensions. (#72999)

khushi-411 committed 2 years ago

905efa82

Remove istft python functional wrapper (#71993)

peterbell10 committed 2 years ago

9ad0578c

[FSDP] s/local_state_dict/_local_state_dict (#73532)

rohan-varma committed 2 years ago

6bb13f77

retry step for Build and upload nightly docker

cdeepali committed 2 years ago

5ae5f00a

Fix alignment, make sure release labels are included (#73739)

atalman committed 2 years ago

fc914346

Support `0`s in `out_size` of `FractionalMaxPoolNd`

Emilio Castillo committed 2 years ago

3186e366

dbr quant: add torchscript pass to remove redundant aliases (#71230)

vkuzo committed 2 years ago

bf896a29

quantization: fix bug in QuantWrapper with DeQuant qconfig (#73671)

vkuzo committed 2 years ago

eb8d0659

[GHA] Increase ASAN test timeout to 5h (#73740)

malfet committed 2 years ago

Verified 71ed3889

[Easy][c10d][DDP] (Reland) Minor fixes (#73569)

awgu committed 2 years ago

a37d54b6

[DDP][BE] (Reland) Remove bucket replicas (#73567)

awgu committed 2 years ago

7cce2d9d

Back out "[pytorch][PR] Support dataclasses in TorchScript"

Apoorva Garg committed 2 years ago

63932edc

(torch/elastic) add documentation clarifying that torchrun is a console script to torch.distributed.run (#73598)

Kiuk Chung committed 2 years ago

1a8bd1a7

Use intrusive_ptr for LazyTensor (#73445)

wconstab committed 2 years ago

b4173b80

[FX] Make module getattr wrapper proxy buffers (#73612)

James Reed committed 2 years ago

dae7ed17

[_shard] use absolute imports for ShardMetadata instead (#73678)

wanchaol committed 2 years ago

ba205a7e

[FSDP] Add the unittests for the _replace_by_prefix (#73530)

fegin committed 2 years ago

e7a786ff

Add note in RPC docs about retries. (#73601)

pritamdamania committed 2 years ago

71aa3ab0

[quant][core] Add Embedding and EmbeddingBag reference module (#73436)

jerryzh168 committed 2 years ago

2ab97029

[PyTorch FX] Return mapping of qualified names from split_module() (#73564)

kwen2501 committed 2 years ago

d14de313

Support CSR to COO conversion in to_sparse(2). (#73642)

pearu committed 2 years ago

4168c87e

[Static Runtime] Introduce StaticNodeInfo to store ProcessedNode's data independent from runtime instances (#73536)

d1jang committed 2 years ago

bbc59ff2

fix wrong indexing of class names in docs

ain-soph committed 2 years ago

e8b10b6e

batch_norm_jvp: improve error message when running_{mean,var} have forward grad defined (#73655)

soulitzer committed 2 years ago

e6afa4f7

setup system includes for generated files on MacOS (#73591)

mikey dagitses committed 2 years ago

715c000c

static_cast value to microsecond type (#73595)

mikey dagitses committed 2 years ago

f087f1a3

only pass --std=c++14 to C++ compilation actions (#73593)

mikey dagitses committed 2 years ago

5dcc2496

[PyTorch] [Model Tracer] Use c10::Synchronized<T> abstraction for mutex protected data (#73408)

dhruvbird committed 2 years ago

26463aa8

Newer Older