strisunshinewentingwang/pytorch

Pull Requests Commits

Add vision onto the commit pins train

ezyang committed 2 years ago

8cd23eb4

Centralize commit pins in a folder

ezyang committed 2 years ago

004b6387

Make FakeTensors return meta within kerenl invocation, add FakeTensor op tests

eellison committed 2 years ago

3c5a3ca9

[Pytorch] Disable use of qnnpack with ceil_mode avgpool (#79028)

kimishpatel committed 2 years ago

7b42ecad

Prevent invariant assert failure on win+debug (#78636)

George Gensure committed 2 years ago

7541b07f

commit hash update - fix error when no change (#79162)

clee2000 committed 2 years ago

9dd8008a

Support both train / eval modes for ModuleInfo

jbschlosser committed 2 years ago

12658fcd

[JIT] Improve error message for InferredType

davidberard98 committed 2 years ago

17862ae8

move substitute lib to open source (#79093)

linbinyu committed 2 years ago

bfaa187f

refactor op handling to use register pattern

eellison committed 2 years ago

d6ecdf16

Add CPU Fallback

eellison committed 2 years ago

fe7a1349

Migrate FakeTensors to always call into FakeTensorMode and have them hold a reference

eellison committed 2 years ago

290d0979

[Test] create shared targets for xplat aten (#78345)

linbinyu committed 2 years ago

430955b3

Add embedding_bag meta functions

ezyang committed 2 years ago

50f2af84

[DDP] Fix broadcast for channels-last tensors (#79060)

eqy committed 2 years ago

7710d872

Add randomness case to the autograd notes

lezcano committed 2 years ago

a8ea58af

Add detailed error message for iOS test (#79140)

linbinyu committed 2 years ago

7cb4a768

Support indexing of the underlying tensors for nested tensors (#78934)

YifanShenSZ committed 2 years ago

6ad51c94

[fix] composite compliance: margin_ranking_loss, hinge_embedding_loss (#78935)

kshitij12345 committed 2 years ago

e85f3b58

[Vulkan] Replace Conv2dOpContext by VulkanOpContext (#78818)

manuelcandales committed 2 years ago

abe55562

[Vulkan] Replace TransposeConvolution2dContext by VulkanOpContext (#78817)

manuelcandales committed 2 years ago

95806fce

[Vulkan] Replace GruOpContext by VulkanOpContext (#78816)

manuelcandales committed 2 years ago

65ecc02a

[Vulkan] Replace LinearOpContext by VulkanOpContext (#78815)

manuelcandales committed 2 years ago

96445b90

Revert "Port `index.Tensor` to structured kernels."

pytorchmergebot committed 2 years ago

4b82ef79

Move test/jit/test_onnx_export.py to test/onnx (#78116)

titaiwangms committed 2 years ago

c19cf34f

Swap fx2trt_oss to torch_tensorrt (#950) (#79115)

Wei Wei committed 2 years ago

79ee0c0d

Port `index.Tensor` to structured kernels.

bdhirsh committed 2 years ago

cfd84125

[quant][core][improvement] Added earlier termination and improved error message for calling min and max ops on per channel quantized tensors.

dzdang committed 2 years ago

b12d951f

[ONNX] Enable script tests based on recent exporter changes (#77254)

shubhambhokare1 committed 2 years ago

d26c575f

Add onnx support for `torch.lerp` (#78891)

qqaatw committed 2 years ago

b45d303d

Newer Older