strisunshinewentingwang/pytorch

Pull Requests Commits

Add missing format string (#93866)

ezyang committed 1 year ago

498c6ed8

[CI] Add Py-3.11 wheels for all platforms (#93400)

malfet committed 1 year ago

87b9ab48

Disable cudagraphs by default (#93253)

jansel committed 1 year ago

2ea3036d

ConfigModule for _{dynamo,inductor}.config (#93252)

jansel committed 1 year ago

45eadc2c

[mta][foreach] Implement fused adamw (#88015)

crcrpar committed 1 year ago

a23ed38f

[pruning][core][feature] LSTM Structured Pruning prune_functions + pattern (#90801)

jcaip committed 1 year ago

86ab4d49

Enable `USE_CUDA` (#92640)

r-barnes committed 1 year ago

f577a527

Move bazel back to pull (#93867)

clee2000 committed 1 year ago

e80af53b

pnp: move shadow loggers to parent module (#91428)

vkuzo committed 1 year ago

6fe234ec

ns: change PNP testing to use QNNPACK (#91421)

vkuzo committed 1 year ago

56f94756

Add retries for get_workflow_job_id and try catch in upload_test_stats (#93401)

clee2000 committed 1 year ago

1dcd2609

Clean up leftover processes on non-ephemeral Windows runner (#93414)

huydhn committed 1 year ago

eb987abd

[docs] Add section about tensor hooks on in-place in autograd note (#93116)

soulitzer committed 1 year ago

77cbaedd

add filelock as a dependency (#91607)

soumith committed 1 year ago

76b99980

fix(fx): make all `make_fx` invocations isolated (opaque to higher `make_fx` invocations) by default (#93290)

jon-chuang committed 1 year ago

d5901fcc

[BE]: Fix CMake LTO policy on pytorch (#93388)

Skylion007 committed 1 year ago

2fc2ca76

[dynamo] getattr for EnumVariables (#93397)

angelayi committed 1 year ago

bf2e2fea

More fixes and improved clang-tidy checkers (#93213)

cyyever committed 1 year ago

37f7c00a

[inductor] only check mutations attr for TritonKernel (#92277)

chunyuan-w committed 1 year ago

679e869a

[fx] add SymPy assumptions to `FloorDiv` (#93185)

nkaretnikov committed 1 year ago

c4ccf7e1

[Re-open 90267] [inductor] weight prepack for single conv_transpose2d (#91956)

chunyuan-w committed 1 year ago

f1030dcc

Use symbolic tracing_mode for aot repro with dynamic_shapes (#93393)

ezyang committed 1 year ago

66fd99cc

use aten parallel on lu factor (#93037)

zhuhaozhe committed 1 year ago

298075e1

cherry-picking autodiff support for gather/index_select (#93333)

jjsjann123 committed 1 year ago

bdca5fcd

_sparse_coo_tensor_with_dims_and_tensors backward: simplify and optimize (#91704)

nikitaved committed 1 year ago

b484d17c

[jit] jit._drop fun modifier to allow in jit class non-jit decl funs (#93012)

IvanKobzarev committed 1 year ago

6a2838ee

sparse_mask: extend lhs to sparse COO tensors (#92248)

nikitaved committed 1 year ago

994f85d6

Introduce core_aten_decompositions (#93131)

SherlockNoMad committed 1 year ago

6a7d6cc3

[Quant] X86 qengine always uses fbgemm kernels on OS other than Linux (#93218)

Xia-Weiwen committed 1 year ago

f77f88fb

Fix test_file_system_checkpoint_cpu.py temp directory usage (#93302)

huydhn committed 1 year ago

776079b5

Newer Older