strisunshinewentingwang/pytorch

Pull Requests Commits

Also disable inplace fw AD for acos on windows (#62360)

albanD committed 3 years ago

945d40dc

Allow FX tracer to trace control flow (if/while) statements when parameter shapes are in the conditionals (#61820)

Jerry Cai committed 3 years ago

1b147a52

Exclude time of waiting in queue from gloo communication prof… (#61342)

guyang3532 committed 3 years ago

4ed88588

Callable activation function support for Transformer modules (Python) (#61355)

jbschlosser committed 3 years ago

35307b13

[DDP] Make compute_bucket_assignment_by_size return per bucket sizes (#62231)

rohan-varma committed 3 years ago

1f2b96e7

[DDP][ez] Remove misleading comment (#62230)

rohan-varma committed 3 years ago

c76daa6d

[DDP] Save bucket size limits (#62229)

rohan-varma committed 3 years ago

842228fd

Provide option to pass module instance to _load_state_dict_pre_hooks. (#62070)

pritamdamania committed 3 years ago

cac4aa71

[Model Averaging] Update model averager API to avoid the redundant `params` arg needed by post-localSGD optimizer (#62132)

Yi Wang committed 3 years ago

2eaf71d7

[Model Averaging] Post-localSGD optimizer (#62131)

Yi Wang committed 3 years ago

55bee449

[DDP] Log unused param names under DETAIL debug mode. (#62209)

rohan-varma committed 3 years ago

58d45d95

Add actionlint (#62364)

driazati committed 3 years ago

24ed6e6b

Split zeta_kernel out of BinaryMiscOpsKernel.cu (#62261)

peterbell10 committed 3 years ago

fcc7fbe1

ns for fx: fix nit in default qlinear weight extraction function (#62334)

vkuzo committed 3 years ago

f6e13759

ns for fx: fix bug for user function in weight extraction (#62333)

vkuzo committed 3 years ago

72c943a2

[pruner] add cuda tests for pruner (#61993)

kazhou committed 3 years ago

d98b1c40

irange-ify 10 (#62122)

r-barnes committed 3 years ago

b39b28ce

irange-ify 6 (#62115)

r-barnes committed 3 years ago

88f8f2ab

irange-ify 11 (#62121)

r-barnes committed 3 years ago

9e77113e

irange-ify 7 (#62117)

r-barnes committed 3 years ago

b5867a1b

Revert D29928698: [pytorch][PR] Use private squid proxy

janeyx99 committed 3 years ago

59bb4f2d

Port `slow_conv_transpose2d` to structured (#55503)

Meghan Lele committed 3 years ago

3a2603bc

[pytorch] Bring back RemoveInplaceOps() (#62200)

Meghan Lele committed 3 years ago

05b802d4

[Static Runtime] Fixed another build failure in OSS due to test_utils.h (#62338)

navahgar committed 3 years ago

b91a9176

ENH Adds no_batch_dim support for pad 2d and 3d (#62183)

thomasjpfan committed 3 years ago

7c588d5d

Use private squid proxy (#62244)

zhouzhuojie committed 3 years ago

6da4a255

[Model Averaging] Create a base class for model averaging (#62111)

Yi Wang committed 3 years ago

2581dfc2

Revert D29794666: Remove faulty process group code

H-Huang committed 3 years ago

a15fff0a

ENH Adds no_batch_dim tests/docs for Maxpool1d & MaxUnpool1d (#62206)

thomasjpfan committed 3 years ago

71a6ef17

[quant][graphmode][fx] Add reference pattern support for BatchNorm (#62215)

jerryzh168 committed 3 years ago

cdf85a82

Newer Older