strisunshinewentingwang/pytorch

Pull Requests Commits

voznesenskym committed 2 years ago

30736f91

voznesenskym committed 2 years ago

1c6992f5

voznesenskym committed 2 years ago

d2933857

Merge commit 'd902f0b2793cf20ee4cad3fa2b70c648eeab081e'

voznesenskym committed 2 years ago

a15a3e17

voznesenskym committed 2 years ago

b97f9c93

voznesenskym committed 2 years ago

d902f0b2

Reland "Fix dynamo handling for tensor attributes: T, H, mT, mH (#90463)" (#91897)

anjali411 committed 2 years ago

c887837e

Torch package support in dynamo (#91821)

voznesenskym committed 2 years ago

3726d232

RM (unused?) has_mutation (#91931)

voznesenskym committed 2 years ago

ae2e755f

[pruning][core][feature] Add in SaliencyPruner to pruner._experimental (#91814)

jcaip committed 2 years ago

32e9b29c

Dynamo.export uses dynamic=True for symbolic tracing (#91899)

SherlockNoMad committed 2 years ago

42a63a7e

fixing test_batch_norm_implicit_dtype_promotion (__main__.TestNvFuserDynamo) (#91541)

jjsjann123 committed 2 years ago

4919e119

Add Skylion007 as a core reviewer (#91890)

ezyang committed 2 years ago

67f965b1

Disable win vs2019 cpu build+test until we figure out the linker crash (#91932)

ZainRizvi committed 2 years ago

b0f359a3

Add support for logaddexp(float16) in CUDA and implement its reference (#91869)

lezcano committed 2 years ago

138a0188

Pin onnx-script to a version before they bumped numpy (#91929)

ZainRizvi committed 2 years ago

df3adbd5

Refactor stack_trace preservation for node meta preservation (#90803)

SherlockNoMad committed 2 years ago

0f1302ee

Add merged label to ghstack prs (#90238)

clee2000 committed 2 years ago

1e768c63

[4/N] Add test for partial training for NamedOptimizer (#91344)

fduwjj committed 2 years ago

32356aae

Reduce #iters to make test run always (#91837)

Michael Gschwind committed 2 years ago

26beb46d

Add log_once to fused attention kernels (#91858)

drisspg committed 2 years ago

95e3e339

Reland "Add torch.utils.device_mode" (#91796)

ezyang committed 2 years ago

333540a4

Foreach clamp_min clamp_max (#91384)

milesial committed 2 years ago

9d20d6d5

Consider updating pin to CMake version 3.23.1 on run_torchbench CI (#91739)

matthewgraca committed 2 years ago

b2646dcb

Enable bfloat16 for hardtanh_backward_cuda (#91511)

cchan committed 2 years ago

d4aa807b

Fix Dynamo+DDP documentation (#91832)

wconstab committed 2 years ago

630ef6c7

Print and zip remaining test logs (#91510)

clee2000 committed 2 years ago

e67f5ab6

[primTorch] Move `logsumexp` decomp to refs (#91860)

nkaretnikov committed 2 years ago

00e5f3a9

Revert "Fix dynamo handling for tensor attributes: T, H, mT, mH (#90463)"

pytorchmergebot committed 2 years ago

84266ae6

Revert "Torch package support in dynamo (#91821)"

pytorchmergebot committed 2 years ago

f6c7cf1b

Older