strisunshinewentingwang/pytorch

Pull Requests Commits

[SDPA] Mirror c++ implementation in FlashAttention meta func (#106477)

drisspg committed 1 year ago

f533791c

Remove unused workflow.py (#106340)

gmagogsfm committed 1 year ago

b3c29cd1

Register ONNX exporter under PT2 logging (#105989)

BowenBao committed 1 year ago

cebc11ae

[FSDP][Easy] Allow `ModuleWrapPolicy` to take `Iterable` (#104999)

awgu committed 1 year ago

640a96df

[FSDP][7/N] Add warning about frozen params (#104967)

awgu committed 1 year ago

031ce0fa

[dynamo] Add missing fields for THPPyInterpreterFrame. (#103227)

zhxchen17 committed 1 year ago

bdcc454b

[FSDP][6/N] Check valid param freezing for `ModuleWrapPolicy` (#104427)

awgu committed 1 year ago

a8c52863

Pin conda to 23.5.2 for Docker builds (#106473)

atalman committed 1 year ago

aec8418b

Fix grad higher order handling TupleVariable (#106425)

ydwu4 committed 1 year ago

1f1dfa9b

AOTInductor compile in prod env (#106442)

hl475 committed 1 year ago

f9988691

[profiler] add profiler parsing support for custom device. (#106142)

MooYeh279 committed 1 year ago

fb6652b5

Update export/export-aot-inductor benchmark code (#106323)

angelayi committed 1 year ago

6339f57f

Add support for edge dialect ops in exir/serde (#106371)

tarun292 committed 1 year ago

3143d81f

Document `f` parameter of `torch.save` (#106248)

JaredLevi18 committed 1 year ago

cc38d40c

Add type annotations to torch/__init__.py (#106214)

r-barnes committed 1 year ago

1534af2a

Replace `sympy.solve` with a new simplified one. (#105877)

ysiraichi committed 1 year ago

bd84651e

[Quant][PT2E] Re-enable test case of conv add/add_relu recipe for x86inductorquantizer (#105638)

leslie-fang-intel committed 1 year ago

bfed2da2

[BE] document more of FSDP checkpointing logic with a sprinkle of cleaning (#106069)

janeyx99 committed 1 year ago

7e47343d

update tf32 thresholds for H100 (#105879)

Fuzzkatt committed 1 year ago

ae1c0f42

Disable TV if Z3 is not found. (#106399)

ysiraichi committed 1 year ago

7820bd84

freezing w aot (#105497)

eellison committed 1 year ago

57f2a8d3

Use ProcessPoolExecutor in the ufmt adapter (#106123)

jkseppan committed 1 year ago

63b7be5a

add pruning method: Filter Pruning via Geometric Median for Deep Convolutional Neural Networks Acceleration (#95689)

ArmandXiao committed 1 year ago

e7b24308

Check that mypy is installed (#106212)

jkseppan committed 1 year ago

aa0b4dac

Revert "Add nn.CircularPad{*}d for consistency + fix no_batch_dim support (#106148)"

pytorchmergebot committed 1 year ago

dfcfd5ce

[ROCm] fix ROCm 5.5 nightly build after hipblas change (#106438)

jeffdaily committed 1 year ago

b37a50af

[core][pruning][feature] cuSPARSELt build integration (#103700)

jcaip committed 1 year ago

f81f9093

Revert "Add error checking for padding modules (#106147)"

pytorchmergebot committed 1 year ago

d83b887f

Revert "faketensor: prevent deepcopy from cloning FakeTensorMode (#104476)"

pytorchmergebot committed 1 year ago

fdd4b3aa

[quant][pt2e][quantizer] Suppoert set_module_type in XNNPACKQuantizer (#106094)

jerryzh168 committed 1 year ago

d528a137

Newer Older