strisunshinewentingwang/pytorch

Pull Requests Commits

Inductor cpp wrapper: change inputs args from tuple to vector (#90754)

chunyuan-w committed 1 year ago

2ba5c1d7

[FSDP][Easy] ufmt files (#90858)

awgu committed 1 year ago

39d9dd13

[CUDA] Use accumulate type to improve accuracy of grid_sample on half precision inputs (#90427)

xwang233 committed 1 year ago

670efb97

[cuDNN][cuDNN V8 API] (re-re-open) cuDNN V8 API on by default (#89022)

eqy committed 1 year ago

eecd621f

[ao] fixing public v private for torch.ao.nn.X (#87883)

HDCharles committed 1 year ago

6a866c3e

Only populate real_value_cache during export (#90468)

ezyang committed 1 year ago

edc5bb5f

[ao][fx] fixing public v private graph_module.py (#88395)

HDCharles committed 1 year ago

f286cbeb

Fix _fix_weakref memory leak (#90823)

ezyang committed 1 year ago

283cf718

add autocast keys to pybind11 DispatchKey object (#90821)

vkuzo committed 1 year ago

d19791e4

Serialize dynamo/inductor config for minifier (#90501)

williamwen42 committed 1 year ago

86269852

[JIT] Frozen Graph Linear-BatchNormNd Folding (#86706)

min-jean-cho committed 1 year ago

e585156c

[ao] quantize.py fixing public v private (#87521)

HDCharles committed 1 year ago

1ca9d43d

[Quant][fx][bc-breaking] Add simpler BackendConfig pattern format (#90698)

andrewor14 committed 1 year ago

691a44f4

Run MPS PR tests on both Ventura and Monterey (#89312)

malfet committed 1 year ago

1e347b73

[DataLoader2] Fix apply_sharding to accept one sharding_filter per branch (#90769)

ejguan committed 1 year ago

7a112c43

[FSDP][BE] Remove `_module_to_handles`, `HandleConfig`; use term "fqn"; clarify docs (#90840)

awgu committed 1 year ago

1ba4e3c7

Add macro C10_AS_INTARRAYREF_SLOW (#90675)

ezyang committed 1 year ago

8090cb53

replace skipIf with xfailif (#90368)

kshitij12345 committed 1 year ago

cdf4a80c

[BE] Tweak Meta copyright headers (#90805)

malfet committed 1 year ago

fb18c294

[torchgen] Introduce Executorch types and signatures (#90781)

larryliu0820 committed 1 year ago

f3393b7e

[torchgen] Let native function declaration generation logic take a callable (#90780)

larryliu0820 committed 1 year ago

4adffe6d

Align max_pool1d Error Checking between CPU and CUDA/CPU requires_grad (#90211)

deandyu committed 1 year ago

df58020b

[MPS] Skip tests broken on Ventura (#90843)

malfet committed 1 year ago

3859aace

Improve interpolate() speed for channels_last CPU videos (#90302)

NicolasHug committed 1 year ago

8a21cac3

Revert "[functorch] Refactor life handle storage (#90317)"

pytorchmergebot committed 1 year ago

0cd69d7c

fix aot autograd for None fw inputs (#89975)

bdhirsh committed 1 year ago

3c637e80

Add torch.compile support to minifier (#90308)

williamwen42 committed 1 year ago

e9dc8cc1

Inductor cpp wrapper: support bmm, mm, addmm extern call (#88667)

chunyuan-w committed 1 year ago

fde5646f

[SDPA] Standardizes the return shape for dense tensor of SDPA regardless of fused kernel called (#90776)

drisspg committed 1 year ago

51c6c5e1

Give linting steps a unique prefix (#90705)

ZainRizvi committed 1 year ago

caa05e6f

Newer Older