strisunshinewentingwang/pytorch

Pull Requests Commits

[jit][static] Replace deepcopy with copy (#43182)

hlu1 committed 4 years ago

25dcc28c

.circleci: Add CUDA 11 to nightly binary builds (#43366)

seemethere committed 4 years ago

51861cc9

Raise error on device mismatch in addmm (#43505)

peterbell10 committed 4 years ago

42f6c3b1

.jenkins: Remove openssh installs (#43597)

seemethere committed 4 years ago

7beeef2c

Fix type annotation errors in torch.functional (#43446)

rgommers committed 4 years ago

573940f8

fix typo in test_dataloader test_multiprocessing_contexts (take 2) (#43588)

malfet committed 4 years ago

2b70f827

Benchmarks: temporarily disable profiling-te configuration. (#43603)

Mikhail Zolotukhin committed 4 years ago

c1553ff9

[TensorExpr] Start using typecheck in the fuser. (#43173)

Mikhail Zolotukhin committed 4 years ago

3ec24f02

[JIT] Subgraph utils: add an optional vmap argument to the API to allow retrieving value mappings. (#43235)

Mikhail Zolotukhin committed 4 years ago

b763666f

[TensorExpr] Fuser: disallow aten::slice nodes. (#43365)

Mikhail Zolotukhin committed 4 years ago

d18566c6

[TensorExpr] Fuser: only require input shapes to be known (output shapes can be inferred). (#43171)

Mikhail Zolotukhin committed 4 years ago

8dc4b415

[TensorExpr] Fuser: move canHandle and some other auxiliary functions into TensorExprFuser class. (#43170)

Mikhail Zolotukhin committed 4 years ago

f6b7c6da

Back out "Make grad point to bucket buffer in DDP to save memory usage" (#43557)

Haoran Li committed 4 years ago

f35e0696

check in intel nnpi 1007 into fbcode/tp2

hl475 committed 4 years ago

58666982

Batching rule for torch.pow, torch.result_type (#43515)

zou3519 committed 4 years ago

b3f88340

Black to Block for various files (#42913)

DeepakVelmurugan committed 4 years ago

c9f125bf

Evenly distribute output grad into all matching inputs for min/max/median (#43519)

zasdfgbnm committed 4 years ago

348e78b0

Revert D23306683: [quant][graphmode][fx] Testing torchvision

Mikhail Zolotukhin committed 4 years ago

be637fd5

Back out D23047144 "[2/3][lite interpreter] add metadata when saving and loading models for mobile"

husthyc committed 4 years ago

05f27b18

Remove unnecessary copies in ProcessGroupGloo for multiple inputs allreduce (#43543)

rohan-varma committed 4 years ago

5ca6cbbd

Correct the windows docs (#43479)

peterjc123 committed 4 years ago

9b05fbd9

Update the QR documentation to include a warning about when the QR.backward is well-defined. (#43547)

nikitaved committed 4 years ago

3df398a3

[quant][graphmode][fx] Testing torchvision (#43526)

jerryzh168 committed 4 years ago

62dcd253

Revert D23299452: [pytorch][PR] fix typo in test_dataloader test_multiprocessing_contexts

zhangguanheng66 committed 4 years ago

9420c773

Polish the nightly.py docs in CONTRIBUTING a little (#43494)

rgommers committed 4 years ago

ebc0fc4d

Grammatical corrections (#43473)

Łukasz Goncerzewicz committed 4 years ago

3dcfe848

Fix typo in LSTMCell document (#43395)

ttyio committed 4 years ago

f32ca57c

Allocating warp to an input index in compute_cuda_kernel (#43354)

kiranmatam committed 4 years ago

f8e9e7ad

move wholearchive to link option (#43485)

mszhanyi committed 4 years ago

76894062

Refactored the duplicate code into a function in _ConvNd (#43525)

noelmat committed 4 years ago

1089ff40

Newer Older