strisunshinewentingwang/pytorch

Pull Requests Commits

review comments

eqy committed 2 years ago

fe7e2aa1

Update select_compute_arch.cmake

eqy committed 2 years ago

3116dfa0

add `sm90` to `cpp_extension.py`

eqy committed 2 years ago

d523b461

enable scalar reduction with dim=-1 (#88628)

Natalia Gimelshein committed 2 years ago

53ca5ad3

Dynamo DDP accuracy bench uses find_unused_parameters (#88645)

wconstab committed 2 years ago

89c58196

Clean up SymFloat binding to cover all functions (#88370)

albanD committed 2 years ago

fcc28834

Fix categorization of sym_int method (#88369)

albanD committed 2 years ago

6abaa594

Add torch.distributed.DistBackendError exception type, thrown from C10D_NCCL_CHECK (#88134)

H-Huang committed 2 years ago

bc66ddb5

Create _make_alias to preserve the name of a function when creating an alias (#88114)

lezcano committed 2 years ago

1a7c4b0d

nvprims bookend non compute (#88457)

jjsjann123 committed 2 years ago

af09270e

Revive static_runtime_benchmark build and test (#87660)

huydhn committed 2 years ago

8cb5c554

[ci] increase timeout time of ios test app build (#88611)

suo committed 2 years ago

02c1a304

[Autograd] Use in-place input accumulation fast path for dense Tensors. (#88339)

Taylor Robie committed 2 years ago

8f66ae41

Add missing args to DDP constructor in distributed.pyi (#88209)

yhcharles committed 2 years ago

ffb6e689

[Pytorch] add an option to disable TORCH_WARN and TORCH_WARN_ONCE log (#87188)

biubiuX committed 2 years ago

ced71e8e

[vision hash update] update the pinned vision hash (#88465)

pytorchmergebot committed 2 years ago

ed97e0aa

Setting pickle_module isn't working (#88570)

boringcrypto committed 2 years ago

9f11ce7f

Add support for symbolic shapes to sparse tensor (#88573)

ezyang committed 2 years ago

825f4e60

[LTC] Remove view (#88445)

alanwaketan committed 2 years ago

c29502dd

Do not use double for single-prec upsample (#88277)

malfet committed 2 years ago

f2000842

Fix typos used in documents under torch directory (#88483)

kiszk committed 2 years ago

4ea2310f

[Reland] Use sudo when reset NVIDIA devices (#88605)

huydhn committed 2 years ago

d25be63c

Implement a constructor for nested_tensor that is similar to torch.tensor() (#88213)

Antoni Viros i Martin committed 2 years ago

c77368d4

Pin linux ninja dep to 1.10.2 (#88548)

huydhn committed 2 years ago

72a73519

Use test/test-reports for inductor (#88533)

huydhn committed 2 years ago

fdf28651

Fix segfault in has_torch_function (#88559)

peterbell10 committed 2 years ago

eb3f975c

Fix pull docs build running with a schedule and increase cpp doc timeout to 4h (#88589)

huydhn committed 2 years ago

4796e23b

Add a note on the stability of linalg functions. (#88313)

lezcano committed 2 years ago

d453b3c4

Revert "fallback for scatter_(scalar) (#88210)"

pytorchmergebot committed 2 years ago

b00c43b3

Dynamo Dashboard Improvements (#88516)

williamwen42 committed 2 years ago

0e67b2f7

Older