strisunshinewentingwang/pytorch

Pull Requests Commits

Revert "[LTC] Make some LazyGraphExecutor private data structures protected (#90457)"

pytorchmergebot committed 1 year ago

b7dfbf87

[fx] Added better tests to pass infra (#90432)

angelayi committed 1 year ago

02eb0bdb

Fix non-existing parameters in docstrings (#90505)

kit1980 committed 1 year ago

f51f6aa3

[inductor] Update TIMM skip list (#90188)

desertfire committed 1 year ago

fd3f5d7b

[FSDP] Subtest `CPUOffload` for `test_fsdp_grad_acc.py` (#90545)

awgu committed 1 year ago

1a735a80

[SDP] Fix alignment check for efficient_attention (#90413)

drisspg committed 1 year ago

912748e3

Use some `if constexpr` in the code (#90483)

malfet committed 1 year ago

669f7461

[reland][dynamo] use optimizers correctly in benchmarking (#87492)

anijain2305 committed 1 year ago

d91d7a32

[dynamo] Add is_compiling for dynamo (#90329)

mlazos committed 1 year ago

9c4189f8

[FSDP] Allow nested FSDP wrapper to use different mixed precision (#90523)

mrshenli committed 1 year ago

08245060

Log1p complex for CUDA (#90422)

mfkasim1 committed 1 year ago

eedf7a49

Revert "[inductor] New approach for computing triton load/store masks (#89566)"

pytorchmergebot committed 1 year ago

b2795d3c

use proper temp directories in test_tensorboard.py (#89826)

ppwwyyxx committed 1 year ago

4e1881b8

Fix: Make `__len__` of datapipes dynamic (#88302)

fgerzer committed 1 year ago

09ccda0d

[LTC] Make some LazyGraphExecutor private data structures protected (#90457)

alanwaketan committed 1 year ago

93aa6e3e

Disable BUILD_CAFFE2 from ONNX builds (#90475)

Thiago Crepaldi committed 1 year ago

bcf7036b

Add logging for aot autograd and unified debug flag (#88987)

mlazos committed 1 year ago

730e44bb

[Vulkan] Enable QInt8 weights and test quantized convolution with QInt8 weights and QInt32 bias (#90441)

manuelcandales committed 1 year ago

983d4f6f

[inductor][Reland] Use decomposition for _to_copy (#90494)

desertfire committed 1 year ago

282dfe8b

Revert "Dynamo, FX, Inductor Progress Bars (#88384)"

pytorchmergebot committed 1 year ago

65810635

Add missing infer_size_symdimvector implementation. (#90405)

ezyang committed 1 year ago

eeb3f8aa

[inductor] New approach for computing triton load/store masks (#89566)

Fabio Rocha committed 1 year ago

c6c2de58

simplify implementation of c10::isIntegralType (#90193)

mikey dagitses committed 1 year ago

c8954a89

Reland "Add heirachical module names to torchFX graph.node" (#90205)

alexmsettle committed 1 year ago

6b7efac3

Implement checks for vmap escaped errors (#89585)

srossross committed 1 year ago

0a008580

[ao] public vs private for ao.quantization._X (#88392)

HDCharles committed 1 year ago

c71b1285

[ao] backend_config moving all to top (#88391)

HDCharles committed 1 year ago

6050a7a3

[threaded PG] fix long hang issue in testing (#90515)

XilunWu committed 1 year ago

3759777e

Dynamo, FX, Inductor Progress Bars (#88384)

msaroufim committed 1 year ago

db0ce4ac

[vision hash update] update the pinned vision hash (#90513)

pytorchmergebot committed 1 year ago

b4c27c86

Newer Older