strisunshinewentingwang/pytorch

Pull Requests Commits

Fake Tensor Docs (#101882)

msaroufim committed 1 year ago

ff5b9428

Add Logging Doc to compile index (#101888)

msaroufim committed 1 year ago

581d13a0

Revert "separate out dynamo .requires_grad and .is_grad_enabled guards (#100570)"

pytorchmergebot committed 1 year ago

7f3fed12

Docs for torchcompile and functorch (#101881)

msaroufim committed 1 year ago

2dd33c71

Update BCEWithLogitsLoss pos_weight description in documentation (#101567)

cviviers committed 1 year ago

81c181dc

match sdpa patterns from HF (#100609)

Natalia Gimelshein committed 1 year ago

5ea7096e

Print restarting analysis at INFO level with a exception breadcrumb (#101573)

ezyang committed 1 year ago

96ee23e1

Update batch size for a couple models (#101837)

eellison committed 1 year ago

e5e451a9

Revert " fix missing-prototypes warnings in torch_cpu (Part 4) (#100849)"

pytorchmergebot committed 1 year ago

498c34e8

Revert "fix inference_mode with torch.compile (#101219)"

pytorchmergebot committed 1 year ago

083f304d

Revert " fix missing-prototypes warnings in torch_cpu (Part 5) (#101788)"

pytorchmergebot committed 1 year ago

e760a968

[export] Error when constraining on static values (#101655)

angelayi committed 1 year ago

4f9aa7cb

[BE]: Enable ruff rule TRY302 and apply fixes (#101874)

Skylion007 committed 1 year ago

3e2ea32d

`collect_env`: parse HIP version exception free (#101844)

malfet committed 1 year ago

1ac663d9

[ONNX] Support aten::broadcast_to (#101833)

titaiwangms committed 1 year ago

0df691df

Fix specify_constraints signature for exporting module (#101831)

ydwu4 committed 1 year ago

113b6705

fix inference_mode with torch.compile (#101219)

bdhirsh committed 1 year ago

11f7ae19

separate out dynamo .requires_grad and .is_grad_enabled guards (#100570)

bdhirsh committed 1 year ago

1fabee39

Check devices on meta functions that return inputs (#101807)

eellison committed 1 year ago

f99eeb5b

inductor: fix FloorDiv issue for dynamic shape path (#101793)

XiaobingSuper committed 1 year ago

61b6b038

fsdp support create hybrid-sharded process group for custom backend (#100622)

medivh-xp committed 1 year ago

e06bd8f3

Add missing conversion functions between half and float for ppc64le (#100168)

vinithakv committed 1 year ago

4441ce21

Improve the functionality of untyped storage for privateuse1. (#100868)

wbigat committed 1 year ago

4486a1d0

SymIntify functorch vmap (#101409)

ezyang committed 1 year ago

f66d5dd7

[reland][opinfo] empty_strided (#101782)

khushi-411 committed 1 year ago

1aaf0396

Fix torchinductor uint8 bug (#101468)

FindHao committed 1 year ago

e5b7c7a0

Support autograd.Function w/ grad (#99483)

voznesenskym committed 1 year ago

4c1bc91f

[ONNX] Detect None constant during jit scalar type analysis (#101608)

Thiago Crepaldi committed 1 year ago

7776a41b

Add cooperative_groups header to cuda_to_hip_mappings.py (#100721)

rraminen committed 1 year ago

eb470ab2

PyTorch -> C++17 (#98209) (#100557)

r-barnes committed 1 year ago

bcb4444c

Newer Older