strisunshinewentingwang/pytorch

Pull Requests Commits

[quant][api] Add assert for backend in get_default_qconfig related apis (#86259) (#87331)

jerryzh168 committed 1 year ago

4caddac5

[FSDP][6/N] Remove FPW! (#87114)

awgu committed 1 year ago

4cc5d664

[FSDP][5/N] Update `FlatParamHandle` after FPW deprecation (#87113)

awgu committed 1 year ago

f8dd2742

[FSDP][4/N] Rework FPW test to not use FPW (#87112)

awgu committed 1 year ago

214d5175

[FSDP][3/N] Register `flat_param` to wrapped module (#87086)

awgu committed 1 year ago

277e37f9

[FSDP][2/N] Remove `_fsdp_wrapped_module.flat_param` (#86122)

awgu committed 1 year ago

9f8ef8ea

Reland "add an API for external backends to register custom device names (#86992)" (#87453)

bdhirsh committed 1 year ago

ce0c6e82

Fix input dimension issue in RNN, LSTM, GRU error message (#87442)

jyx-su committed 1 year ago

70c46d32

Revert "Back out "Revert D40198461: [pytorch][PR] Backport currently dont work with some models if:" (#87124)"

pytorchmergebot committed 1 year ago

0c1dec37

Audit for error prone isinstance int/float and add lint (#87345)

ezyang committed 1 year ago

d73d4aa7

OpInfo: Add test that sample_inputs_func returns a generator (#84567)

peterbell10 committed 1 year ago

1285542f

Reenable `isinstance` with `torch.distributed.ReduceOp` (#87303)

crcrpar committed 1 year ago

aa8248cc

Make LazyGraphExecutor extensible (#87218)

antoniojkim committed 1 year ago

d37dc6f6

Fix typo under torch directory (#87274)

kiszk committed 1 year ago

d80a5f9a

[MPS] Revamp copy_to_mps_ implementation (#86956)

malfet committed 1 year ago

ae62cf7c

[dynamo] [easy] RM spurious `)` (#87439)

voznesenskym committed 1 year ago

435e78e5

Python binding for dispatcher getAllOpNames (#87422)

SherlockNoMad committed 1 year ago

ab901b48

[inductor] Fix channels_last conv2d propagation when CuDNN is not found (#87266)

soumith committed 1 year ago

7caeac17

Fix registration hooks (#87369)

antoniojkim committed 1 year ago

6b59d9b5

[AOT][CUDAGraphs] torchdynamo -> torch._dynamo (#87243)

soumith committed 1 year ago

ff43288d

[functorch] fix AOTAutograd tutorial (#87415)

zou3519 committed 1 year ago

13ab8193

Enable inductor CI for huggingface (#86792)

desertfire committed 1 year ago

b1cf377c

[Inductor] Convert 0d CPU tensor to scalar during triton codegen (#87329)

yanboliang committed 1 year ago

9ba63225

Add `weights_only` option to `torch.load` (#86812)

malfet committed 1 year ago

961ebca2

Remove jansel/voz from dynamo CODEOWNERS (#87430)

jansel committed 1 year ago

e3d73bbb

Improve the performance of validate_non_overlapping_shards_metadata (#85639)

fegin committed 1 year ago

bd1e95ce

Back out "Revert D40198461: [pytorch][PR] Backport currently dont work with some models if:" (#87124)

qihqi committed 1 year ago

a42fbfa0

Revert "[dynamo] use optimizers correctly in benchmarking (#87311)"

pytorchmergebot committed 1 year ago

f38a88c4

[Inductor] TorchInductor tracing fx_graph.py should import overrides (#87271)

yanboliang committed 1 year ago

a91abedf

set ci in mps (#87325)

clee2000 committed 1 year ago

1801b57c

Newer Older