strisunshinewentingwang/pytorch

Pull Requests Commits

Reenable slow gradcheck and make it pass (#80514)

soulitzer committed 2 years ago

f595467e

Update actionlint (#81922)

ZainRizvi committed 2 years ago

d28e6671

Revert "Support non-standard bools in CUDA mode (#79393)"

pytorchmergebot committed 2 years ago

b2f09ad1

Support non-standard bools in CUDA mode (#79393)

peterbell10 committed 2 years ago

7592d3f6

Fix cuda-mode and add more tests (#81898)

peterbell10 committed 2 years ago

1d88635b

Revert "Enable reentrant dispatch for decompositions (#81598)"

pytorchmergebot committed 2 years ago

521d5ae1

Enabling SymInt in autograd; take 3 (#81145)

Krovatkin committed 2 years ago

032facd6

[T124977342] Clean up duplicated code in vulkan_api_test (#81834)

Wayne Chiang committed 2 years ago

18e2aeae

Small optimisation to linalg.cholesky (#81316)

lezcano committed 2 years ago

31c3dcb2

[PrimTorch] Reference for linalg.norm (#81241)

lezcano committed 2 years ago

96dfee4c

[PrimTorch] Reference for linalg.matrix_norm (#81113)

lezcano committed 2 years ago

c5330183

Add shape fn for hardswish and backward (#81827)

JackCaoG committed 2 years ago

5fa64ed9

Cleaning Indexing.cu (#81477)

r-barnes committed 2 years ago

3574329a

remove backend keys from FunctionalTensorWrapper, update TensorImpl::is_device methods (#81471)

bdhirsh committed 2 years ago

ec77d35b

Registered python meta functions to a table (#81092)

Chillee committed 2 years ago

2529ff4b

Add warning about DCE in FX being unsound with mutation (#81818)

ezyang committed 2 years ago

031ec663

Add basic support for sparse meta tensors (#81800)

ezyang committed 2 years ago

44193f6b

Generalize sparse construction to work with arbitrary device types (#81796)

ezyang committed 2 years ago

104cbbf1

Add sparse, quantized and nested tensor meta support to codegen (#81793)

ezyang committed 2 years ago

6f0c2539

Get TensorOptions.h using the FORALL macros to reduce manual code (#81789)

ezyang committed 2 years ago

a33c5ace

Sync up DispatchKey in model with C++ (#81770)

ezyang committed 2 years ago

ec91f42d

Refactor functionality and backend keys to reduce duplication (#81752)

ezyang committed 2 years ago

1724e9f2

Support non-standard bools in CUDA unique (#79392)

peterbell10 committed 2 years ago

635f670d

[functorch] remove kl_div skips (pytorch/functorch#975)

Samantha Andow committed 2 years ago

43523f46

[functorch] Use python3 in shebang

zou3519 committed 2 years ago

c6600e25

[functorch] Add exhaustive testing of vmap autograd composability (pytorch/functorch#851)

Samantha Andow committed 2 years ago

e2361bcc

[functorch] disable tests to land symint autograd changes (pytorch/functorch#973)

Krovatkin committed 2 years ago

cac86626

[functorch] [batch-rule] householder_product (pytorch/functorch#972)

zou3519 committed 2 years ago

aa94afa2

[functorch] Use fake tensor for primal computation in AOTAutograd (pytorch/functorch#935)

ezyang committed 2 years ago

cc07aaf3

[functorch] Allow batch norm with all variations of batching when training=False (pytorch/functorch#958)

Samantha Andow committed 2 years ago

cb627889

Newer Older