strisunshinewentingwang/pytorch

Pull Requests Commits

Rebased dynamic shape POC

Chillee committed 2 years ago

05829fa6

CPUBlas: Use opmath_type for alpha/beta scalars (#65839)

peterbell10 committed 2 years ago

7705175f

Automated submodule update: kineto (#79925)

facebook-github-bot committed 2 years ago

cc0f1cc3

[Static Runtime] Fix precision error in test cases (#80935)

akshayParashar1995 committed 2 years ago

38169c22

Revert "[MPS] Add argmin (#80828)"

pytorchmergebot committed 2 years ago

25564998

[JIT] Document runtime flags (#80900)

davidberard98 committed 2 years ago

46aecea5

[MPS] Add huber loss (#80163)

qqaatw committed 2 years ago

c48e964a

[MPS] Add argmin (#80828)

qqaatw committed 2 years ago

37ebcc8a

Disallow creating a library with prim namespace (#80913)

anjali411 committed 2 years ago

7f37b1b3

fix overload ambiguity with functional ops; fix _foreach op grouping (#80556)

bdhirsh committed 2 years ago

960758b0

fx quant: fix warning in util function when cloning tensors (#80883)

vkuzo committed 2 years ago

ce0786ad

Make addcmul and addcdiv support different dtypes (#74234)

Nikita Karetnikov committed 2 years ago

12382f0a

Add refs::{isposinf, isneginf} (#78655)

zasdfgbnm committed 2 years ago

736fb7d2

[MPS] Add softplus backward (#79873)

qqaatw committed 2 years ago

8745118c

Revert "[jiterator] Reduce templating in jitted_gpu_kernel_impl (#80103)"

pytorchmergebot committed 2 years ago

25e5a3e6

Revert "[jiterator] De-template launch_jitted_reduce_kernel (#80138)"

pytorchmergebot committed 2 years ago

b46f0357

[ONNX] Moved uninitialized_optional test cases into runtime test and parameterized left test cases (#80145)

titaiwangms committed 2 years ago

f608db5f

[MPS] Move the View ops to a separate file and reduce the number of graphs created (#80491)

kulinseth committed 2 years ago

5436134f

Enable ROCm CI for trunk (#80920)

jithunnair-amd committed 2 years ago

6305a85c

Revert "[FSDP Optim State] Remove checkpoint prefix (#80480)"

pytorchmergebot committed 2 years ago

fe361ded

[build] fix compiling with clang13 (#80916)

suo committed 2 years ago

b349d159

fix the invalid configuration argument error when running layer norm backward (#80893)

jianyuh committed 2 years ago

45f77189

Add check for cuda lazy init (#80912)

malfet committed 2 years ago

1ad7ef3f

[FSDP Optim State] Remove checkpoint prefix (#80480)

rohan-varma committed 2 years ago

04c50fec

[Static Runtime] test case for staticRuntime::runAsync() API (#80407)

akshayParashar1995 committed 2 years ago

fefdad61

Move serialize_module() out of OSS graph_manipulation.py to internal (#80785)

dborkovic committed 2 years ago

9402219a

improve sort multi-core perf by adjusting grain_size w.r.t. dim_size (#74897)

mingfeima committed 2 years ago

6d03ea65

Prevent potential memory leaks in CoreML backend (#79928)

jmdetloff committed 2 years ago

f36a5d23

Back out "[profiling] Adding targets file for test_mobile_profiler" (#80789)

bochko committed 2 years ago

b603860c

[torch.fx] Check node type before fetching .users (#80166)

jiaxuzhu92 committed 2 years ago

280f4704

Older