strisunshinewentingwang/pytorch

Pull Requests Commits

Vectorized horizontal flip implementation (#88989)

vfdev-5 committed 1 year ago

7ec8a4d2

[Dynamo] Support Tensor.nelement & torch.cuda.is_available (#89164)

yanboliang committed 1 year ago

81a4aeab

Added partial decomposition of conv_backward and grad_bias computation (#89128)

Chillee committed 1 year ago

8a419cbf

[quant][fx][be] Refactor replace observer with q/dq op code (#89247)

jerryzh168 committed 1 year ago

38ccd08f

Use standard __func__ macro in symbolic shape. (#89264)

zhxchen17 committed 1 year ago

c219b55b

[xplat] remove -weak_framework (#89233)

rmaz committed 1 year ago

12a97444

[Quant] Allow setting fixed qparams for inner LSTM ops (#88456)

andrewor14 committed 1 year ago

19e66fce

[inductor] Skip DALLE2_pytorch in torchbench (#89288)

desertfire committed 1 year ago

19fcb805

[inductor] Temporarily disable functorch_dp_cifar10 test in TorchBench (#89281)

desertfire committed 1 year ago

1f7c0ff6

Update torch.distributed.DistBackendError type (#89235)

H-Huang committed 1 year ago

55e55d95

Add most in-place references/decompositions (#88117)

lezcano committed 1 year ago

154e58c0

Simplify maybe_resize_out (#88116)

lezcano committed 1 year ago

6741443c

Fix names of some reference functions (#88115)

lezcano committed 1 year ago

ce0e22a8

Add platform markers for linux only extra_install_requires (#88826)

JacobHayes committed 1 year ago

2e358cc9

Export c10/[macros|util] headers to be used by internal inductor builds (#89249)

malfet committed 1 year ago

5654fed2

[PT-D][Checkpoint] Update import and update docstring for distributed checkpoint (#89256)

wz337 committed 1 year ago

4c672498

[LTC] Upstream short_metrics (#89186)

alanwaketan committed 1 year ago

2dcacc6b

[ao] maintain BC for is_activation_post_process (#89260)

HDCharles committed 1 year ago

c5fafb4e

Disable tracing `zero_grad()` (#88731)

mlazos committed 1 year ago

30c3e5af

Gate CUDA-only inductor tests by HAS_CUDA (#89251)

huydhn committed 1 year ago

afdc48f8

[flaky] relax tolerance conv1d_vs_scipy (#89193)

kshitij12345 committed 1 year ago

6a964c16

Add support trace on MPS backend (#87910)

pumetu committed 1 year ago

fc1c0cd3

[xnnpack][executorch] remove unordered_set from xnn_compiler (#89231)

mcr229 committed 1 year ago

7beb1518

Always retry curl downloads (#89157)

ZainRizvi committed 1 year ago

ab75982d

Fix consistentcy of histc on CPU and CUDA (#87832)

Aidyn-A committed 1 year ago

3bc78295

Symintify repeat_interleave.self_int (#89111)

SherlockNoMad committed 1 year ago

f1fb586b

Fix tol for test_nvfuser_correctness__softmax_backward_data_cuda (#89178)

SherlockNoMad committed 1 year ago

ba5e39e1

docs: conv2d `padding` attribute- add `int` option (#85004)

YoniChechik committed 1 year ago

6f609dd0

[Executorch] [Quantization] New pattern for dynamic dequant (#89236)

JacobSzwejbka committed 1 year ago

6f4f69f5

[quant][be] Move some helper functions to the top level to reduce function length (#89246)

jerryzh168 committed 1 year ago

f4efc5e8

Newer Older