strisunshinewentingwang/pytorch

Pull Requests Commits

[ci] custom stale action

suo committed 2 years ago

692f1b9a

Add test to make sure submodule hooks fire

rohan-varma committed 2 years ago

a4126e59

[JIT] Prevent nvfuser registration on ROCm

davidberard98 committed 2 years ago

a305c078

Add checks for public and private API

anjali411 committed 2 years ago

31ed4827

[FSDP] Add `ignored_modules` ctor arg

awgu committed 2 years ago

648823b0

Turn on -Wsign-compare

ezyang committed 2 years ago

c2124f5c

Adding opinfo tests for binary cross entropy with logits

drisspg committed 2 years ago

f5f64526

prevent overriding shuffle settings in DataLoader for datapipes

pmeier committed 2 years ago

04db1b87

Turn on sanitize-address-use-after-scope/detect_stack_use_after_retur…

ezyang committed 2 years ago

58f5fd9d

Revert "Extend sign-compare warnings to gcc"

pytorchmergebot committed 2 years ago

80e05b7d

[quant][fx] Use native backend_config_dict in convert

jerryzh168 committed 2 years ago

761bb062

[caffe2] remove unecessary RCCL dependency

minsii committed 2 years ago

42b4d0e9

Extend sign-compare warnings to gcc

malfet committed 2 years ago

34446653

modify the check condition of Conv-> Add/Sub/Mul/Div folding

jiayisunx committed 2 years ago

71587e05

[quant][fx] Use native backend_config_dict in prepare

jerryzh168 committed 2 years ago

f83d0473

Revert "Add checks for public and private API"

pytorchmergebot committed 2 years ago

af920386

[FSDP] Implement reshard_flatten_tensor

fegin committed 2 years ago

577c9ff8

[ROCm] libtorch nightly now correctly uses rocm runners

jeffdaily committed 2 years ago

d38d950f

[quant][core][improvement] Added support for data_ptr<T> for quantized tensors to return pointer to underlying int type (e.g., int8* instead of qint*)

dzdang committed 2 years ago

b1310b46

Add checks for public and private API

anjali411 committed 2 years ago

d7e23286

Adding Caching of calculated Symbolic Shapes

Gamrix committed 2 years ago

9a7bfaa9

[shard] Add basic math ops to ShardedTensor and add ReplicatedTensor inter-op

wanchaol committed 2 years ago

be354d81

patching clamp for one sided clamp

jjsjann123 committed 2 years ago

0203341b

[FSDP] Avoid using the legacy _sharded_tensor package

fegin committed 2 years ago

c2d4e201

disabling reshape

jjsjann123 committed 2 years ago

f7e7af80

updated cudnn frontend to v0.6.1

dzdang committed 2 years ago

25aa251f

[FSDP] summon full params staticmethod

rohan-varma committed 2 years ago

143f7cca

Add note about testing inconsistency

janeyx99 committed 2 years ago

8d3e3ebc

Update docker-builds to add CUDA 11.6

atalman committed 2 years ago

6d9210be

Refractor flatbuffer jit code

pavithranrao committed 2 years ago

6402e624

Newer Older