strisunshinewentingwang/pytorch

Pull Requests Commits

[third-party] Update nccl to 2.9.8 (#58667)

malfet committed 3 years ago

d8412142

Add missing .to_sparse(ndim) gradient (#58413)

kylevedder committed 3 years ago

bbf92e61

Enable `ceil`, `floor`, `frac`, `round` & `trunc` for BFloat16 on CUDA (#57910)

imaginary-person committed 3 years ago

8a3d9962

[jit] Implement ScriptProfile to collect instruction profiles. (#57397)

zhxchen17 committed 3 years ago

034a238b

Adds grid_sampler to autocast fp32 list for 1.9 (#58679)

mcarilli committed 3 years ago

e8c6a650

Do not use TF32 matmul in linalg and DDP tests (#56114)

xwang233 committed 3 years ago

691c1391

Added statistic related to out variant nodes

Edvard Ghazaryan committed 3 years ago

a7f06e1e

turn off deadline for adagrad test

ngimel committed 3 years ago

056287ae

Revert "Striding for lists Part 2 (#49352)" (#58523)

tugsbayasgalan committed 3 years ago

9db64e6e

Cleanup functional.py after lu_unpack was removed (#58669)

heitorschueroff committed 3 years ago

91232296

[nnc] Use int64 to compute matmul flops heuristic (#58676)

bertmaher committed 3 years ago

0e1bed36

Remove opinfo warning from floor_divide (#58682)

Chillee committed 3 years ago

a60ce98a

[Static Runtime] Check input container type in aten::__getitem__ (#58639)

Hao Lu committed 3 years ago

1981904c

.github: Upload /download large artifacts to s3 (#58506)

seemethere committed 3 years ago

84500d03

ENH Adds check for input sizes in cosine_similarity (#58559)

thomasjpfan committed 3 years ago

151ec563

Add Deploy to PredictorContainer (#58503)

wconstab committed 3 years ago

3c55db80

Abladawood patch 1 (#58496)

abladawood committed 3 years ago

1fc3e1e1

masked_scatter thrust->cub (#56750)

zasdfgbnm committed 3 years ago

5152cf86

[DataLoader] Introduce MapMapDataPipe functional datapipe (#58258)

MarcioPorto committed 3 years ago

4942fe02

[DDP] Support not all outputs used in loss calculation (#57081)

rohan-varma committed 3 years ago

faa7d379

Fix dtype inference in sparse_csr_tensor_ctor (#58631)

malfet committed 3 years ago

abb215e2

Fix bug in test_fx_experimental codegen (#58587)

heitorschueroff committed 3 years ago

9ac0bd23

Enables builds with Compute Library backend for oneDNN (#55913)

nSircombe committed 3 years ago

bf00d26d

DOC Adds code comment to clarify nn.Linear.reset_parameters (#58487)

thomasjpfan committed 3 years ago

145a6f79

[pkg] Catch exceptions where dependency resolution gets invalid imports (#58573)

suo committed 3 years ago

5caccbe3

[pkg] simplifications to broken dependency handling (#58572)

suo committed 3 years ago

703f2439

Quote in setup-ci-env (#58637)

malfet committed 3 years ago

c4f0c5ee

Fix GIL issue when acquiring multiple sessions. (#58584)

Ailing Zhang committed 3 years ago

8615fd65

Make torch::deploy work with or without cuda (#58493)

wconstab committed 3 years ago

24786bd6

port `sgn` to structured (#58197)

eqy committed 3 years ago

fbc235c2

Newer Older