strisunshinewentingwang/pytorch

Pull Requests Commits

Add Skip List for Aten Ops that are fused in nvFuser. (#86101)

kevinstephano committed 2 years ago

b14f1d7b

Xformer SDP forward/backward kernel (#86157)

drisspg committed 2 years ago

c5a48440

[vision hash update] update the pinned vision hash (#86173)

pytorchmergebot committed 2 years ago

ca39e367

Fix SubgraphMatcher for case of no anchor found (#86421)

SherlockNoMad committed 2 years ago

2fec853c

Fix cond tests after CI was disabled for a bit (#86321)

voznesenskym committed 2 years ago

b73f0e98

Fix broadcasting to implicit leading dimensions in `torch.where` on MPS (#86240)

aslvrstn committed 2 years ago

ca69ddb4

[refactor] Renaming ao.sparsity to ao.pruning (#84867)

z-a-f committed 2 years ago

0e30da3f

Cleanup PT-D imports (#85781)

dstaay-fb committed 2 years ago

9a170b24

[nll_loss] Avoid unnecessary type casts (#86086)

crcrpar committed 2 years ago

a2419638

Replacement is irrelevant for 1-sample multinomial (#86342)

malfet committed 2 years ago

2232db7f

Declare public dependencies on libshm (#82694)

peterbell10 committed 2 years ago

5a8b07de

Merge more symbolic meta kernels and symint changes from branch (#86334)

bdhirsh committed 2 years ago

08e3999f

Release 1.13: Bump nightly version 1.13->1.14 (#86296)

atalman committed 2 years ago

3af0eafe

Fix SparseAdam consuming iterator (#86210)

ssnl committed 2 years ago

5ed75ec1

[FSDP] Doc to explain running submodules (#86343)

rohan-varma committed 2 years ago

f0977c46

[FSDP] Fix clip_grad_norm for CPU offload (#86337)

rohan-varma committed 2 years ago

3db8ddca

[FSDP] assert to runtime error (#86336)

rohan-varma committed 2 years ago

adfd8f38

CheckpointSequential support non-reentrant (#86331)

rohan-varma committed 2 years ago

7a411952

Docs: fix typo (#86273)

cherrywoods committed 2 years ago

3037f3d7

Revert "Fix memory leak in _LRScheduler.step() (#85602)"

pytorchmergebot committed 2 years ago

233d6f19

Fix for binary upload step, use bash shell rather then default sh (#86382)

atalman committed 2 years ago

bf746798

[ao] fixing public v private for qconfig.py (#86026)

HDCharles committed 2 years ago

facf210f

[kineto] guard global observer init against Edge profiler (#86347)

chaekit committed 2 years ago

7c5e07f8

[torch.ao.quantization] include torch.qint32 for static quant (#86345)

jiaxuzhu92 committed 2 years ago

bc919ac7

Two small improvements to references (#86371)

lezcano committed 2 years ago

08780229

Add total GPU memory utilization (#86250)

huydhn committed 2 years ago

795906f2

Make mergebot message clearer when starting a new merge (#86311)

ZainRizvi committed 2 years ago

1059d3b5

Enable autograd on Linear with sparse COO weight (#86302)

pearu committed 2 years ago

6b295cd0

Support autograd on sparse_mm in full. (#86301)

pearu committed 2 years ago

8f2c2167

Support sum on a sparse COO tensor. (#86300)

pearu committed 2 years ago

88b882cd

Newer Older