strisunshinewentingwang/pytorch

Pull Requests Commits

Add error inputs to `gaussian_nll_loss` `OpInfo` (#88486)

nkaretnikov committed 1 year ago

bbaa0637

Upstream apply_optim_in_backward from TorchRec (#87397) (#88539)

rohan-varma committed 1 year ago

404f254e

Reenable optimizer overlap tests (#88439)

rohan-varma committed 1 year ago

da452bca

Handle case when candidate is empty (#88359)

ezyang committed 1 year ago

d1ee0730

[Reland] Fix primTorch compute_elementwise_output_strides (#88525)

SherlockNoMad committed 1 year ago

46730aec

Functionalize and compute joint simultaneously. (#88063)

ezyang committed 1 year ago

0e3031f7

fx.replace_pattern accepts pattern/replacement as GraphModule (#88479)

SherlockNoMad committed 1 year ago

957a9b63

Add docstring to DDPOptimizer (#88521)

wconstab committed 1 year ago

4bb5c2c2

Add single-process DDP accuracy support to dynamo benchmark suite (#88511)

wconstab committed 1 year ago

1f32c3c0

DDPOptimizer replace debug=True/False with using torchdynamo logger (#88480)

wconstab committed 1 year ago

3fd0729b

nvprims native batch norm patch (#88455)

jjsjann123 committed 1 year ago

52375a0f

[Dynamo] Improve BuiltinVariable log when incorrect arg count happens (#88409)

yanboliang committed 1 year ago

b1116a51

Fix minifier accuracy msg (#88515)

mlazos committed 1 year ago

5220d07d

Populate self.export in InstructionTranslatorBase (#88508)

mergennachin committed 1 year ago

dde9affe

[QNNPACK] Add unaligned attributes where asan fails (#88276)

digantdesai committed 1 year ago

afdc2283

[Quant] Respect non_leaf_module_list for activation modules (#88498)

andrewor14 committed 1 year ago

7560a7b2

[BE] Update native_functions.yaml README; we do not support Tensor! (#88513)

janeyx99 committed 1 year ago

5af3feef

Support DDP ignored parameters in DDPOptimizer (#88460)

wconstab committed 1 year ago

678d0380

enable backward for log1p (sparse layouts) (#88155)

amjames committed 1 year ago

ff6770a9

Support sparse inputs to deg2rad (#88156)

amjames committed 1 year ago

6938dd0b

Enable sparse_csr autograd testing for relu (#88154)

amjames committed 1 year ago

1964d8c3

Add sparse layout support for torch.frac (#88153)

amjames committed 1 year ago

f03302ba

Disable mem leak check (#88373)

clee2000 committed 1 year ago

d632d94c

Re-enable inductor models tests as periodical jobs (#88509)

huydhn committed 1 year ago

093e2208

Don't print fatal:... in generate_torch_version.py (#88335)

janeyx99 committed 1 year ago

3e6579b8

[inductor] Handle the case where kwargs contains tensor (#88417)

desertfire committed 1 year ago

955cbe61

Add nondeterministic error for `scatter` (#88244)

kurtamohler committed 1 year ago

e940a2f8

[fx2ait] fixes for AITSplitter (#87805)

mortzur committed 1 year ago

6575174d

[NVFuser] Upstream push 1026 (#87779)

jjsjann123 committed 1 year ago

7b419e85

[MPS] Fix embedding backward with scalar index (#82809)

qqaatw committed 1 year ago

15e54293

Newer Older