strisunshinewentingwang/pytorch

Pull Requests Commits

Skip superfluous storage allocations while constructing meta tensors (#65331)

cbalioglu committed 2 years ago

6942fccf

[tools_common] Don't remove underscores from call_module targets in get_acc_ops_name (#72664)

jfix71 committed 2 years ago

4737ae7a

Convert type comments to annotations in torch/nn (#72662)

stroxler committed 2 years ago

33b7e6ff

Automated submodule update: FBGEMM (#72665)

facebook-github-bot committed 2 years ago

e4629c60

Don't do BC check on ops with valid upgraders (#72313)

tugsbayasgalan committed 2 years ago

5f2eed61

[FSDP] Add no_sync() context manager (#72446)

awgu committed 2 years ago

426f50e5

Revert D34154832: [pytorch][PR] Add `multi_head_attention_forward` to functional rst docs

Mike Ruberry committed 2 years ago

2fa34fb7

[ATen][rocm] fix type issue (#72676)

shintaro-iwasaki committed 2 years ago

4dae85c3

Add `multi_head_attention_forward` to functional rst docs (#72675)

ProGamerGov committed 2 years ago

bafaf0d6

[TensorExpr] Delet `DimArg` class. (#72390)

Mikhail Zolotukhin committed 2 years ago

1855b149

[TensorExpr] Switch from `ExprPtr` to `ExprHandle` in Compute impl. (#72389)

Mikhail Zolotukhin committed 2 years ago

9123e9b3

Convert type comments to annotations in caffe2/test/onnx/ (#72632)

stroxler committed 2 years ago

730fef25

[ROCM] Add ROCM version api within cmake (#69481)

dllehr-amd committed 2 years ago

eb4e6ca3

pad_sequence: fix regression - support tensor (#72436)

kshitij12345 committed 2 years ago

a2e545e6

ci: Migrate macOS x86_64 binary builds to GHA (#71888)

seemethere committed 2 years ago

c963dedc

Revert D33353157: add native quantization support for pixel_shuffle

suo committed 2 years ago

a34d2849

[SR] Generalize VarStackNodeWrapper (#71573)

Mike Iovine committed 2 years ago

6f84c5f0

improve CPU performance for log_softmax when dim != -1 on both float32 and bfloat16 (#72163)

mingfeima committed 2 years ago

b05d8b95

optimize dim reduce performance on norm, argmax and argmin (#72083)

mingfeima committed 2 years ago

75e76944

Remove fx2trt from oss CI (#72595)

wushirong committed 2 years ago

4d01789f

[SR][easy] CPU fuser uses native control flow (#72544)

Mike Iovine committed 2 years ago

c975b928

[tensorexpr] support for fusing autocasting ops (#72478)

davidberard98 committed 2 years ago

2e042957

Automated submodule update: FBGEMM (#72658)

facebook-github-bot committed 2 years ago

dbbdabc3

[DataPipe] Add docstrings for IterDataPipe and MapDataPipe, along with small doc changes for consistency (#72618)

NivekT committed 2 years ago

3e1eff9a

Make test_multiprocessing_spawn.py compatible with pytest (#50408)

zasdfgbnm committed 2 years ago

634427d6

add native quantization support for pixel_shuffle (#68328)

mingfeima committed 2 years ago

0f8a1fc9

[DataPipe] Adding a note about FileLister behavior (#72619)

NivekT committed 2 years ago

b0dd2c2e

Automated submodule update: FBGEMM (#72647)

facebook-github-bot committed 2 years ago

364055b2

small optimizations of a view function (#72626)

ngimel committed 2 years ago

d891b730

Fix tagged build detection for binary builds (#72628)

malfet committed 2 years ago

b6df02bb

Newer Older