strisunshinewentingwang/pytorch

Pull Requests Commits

TorchDynamo: enable conv+silu fusion (#89278)

XiaobingSuper committed 1 year ago

31708a73

Redefine the simdlen semantic (#89263)

EikanWang committed 1 year ago

bc716383

TorchDynamo: enable conv+relu6 fusion (#89265)

XiaobingSuper committed 1 year ago

79770d36

[Easy] Use prepend arg to register forward hooks in quantize.py (#89391)

mrshenli committed 1 year ago

e0251de4

[vision hash update] update the pinned vision hash (#89287)

pytorchmergebot committed 1 year ago

1db5ce09

use std/libdevice erf in inductor (#89388)

ngimel committed 1 year ago

51e961dd

Temporary increase ASAN shard 5 to 4xlarge (#89387)

huydhn committed 1 year ago

1856fa5d

Revert "Update sdp dispatch logic to enable fused backward (#89154)"

pytorchmergebot committed 1 year ago

e1d58b19

Also include MKL_THREAD_LIB in link libraries for caffe2::mkl (#89378)

ezyang committed 1 year ago

c0992965

Set INTERFACE_LINK_DIRECTORIES on caffe2::mkl (#89359)

ezyang committed 1 year ago

7b0d577c

Fix cat striding in PrimTorch (#89332)

ezyang committed 1 year ago

dbeacf11

Symintify numel(), infer_size, prims.elementwise_meta (#88956)

SherlockNoMad committed 1 year ago

caf3d531

Add support for dynamic kwarg to torch._dynamo.optimize (#89290)

ezyang committed 1 year ago

7c811efa

Revert "Symintify numel(), infer_size, prims.elementwise_meta (#88956)"

pytorchmergebot committed 1 year ago

8ad39536

Add nullptr_t overload to c10::intrusive_ptr (#89196)

kvathupo committed 1 year ago

8ac58bc2

Reland 2 "Towards unifying symbolic and non symbolic fake tensor (#89038) (#89143)" (#89346)

ezyang committed 1 year ago

5582001b

[PT-D][1/N] Sync TP Beta change to prod (#89242)

fduwjj committed 1 year ago

6afe3412

RM expectedFailure UnspecReproTests.test_batch_norm_act_unspec (#89340)

voznesenskym committed 1 year ago

6b8c1b19

[ONNX] Add setType from user into InferredType and Reliable in ConstantValueMap (#88622)

titaiwangms committed 1 year ago

6daf60be

[quant][fix] Add quant_min/quant_max for default dynamic quantization observer (#89267)

jerryzh168 committed 1 year ago

940959eb

Fix try/except flow where DataDependentOutputException is getting wrapped in a RuntimeError (#89314)

voznesenskym committed 1 year ago

808bdbab

Added utility to count memory reads/written in Inductor (#89203)

Chillee committed 1 year ago

419ef2cd

add jvp test with non-contig inputs (#89131)

kshitij12345 committed 1 year ago

7a2930b3

Add --explain flag to bench (#89316)

voznesenskym committed 1 year ago

631baecb

Don't redefine __STDC_FORMAT_MACROS (#89310)

ppwwyyxx committed 1 year ago

e6996ea1

cast C++ py-bound SymNode to SymInt correctly (#89295)

Krovatkin committed 1 year ago

8c0515db

Update sdp dispatch logic to enable fused backward (#89154)

drisspg committed 1 year ago

2e72ec79

[dynamo] mutable local caching to make dynamo faster at tracing mutation (#89170)

mlazos committed 1 year ago

85a87e63

Move bazel to c++17 (#89297)

malfet committed 1 year ago

ea58955d

[dashboard][huggingface] skip accuracy checks for really large models… (#89273)

anijain2305 committed 1 year ago

cad5772c

Newer Older