strisunshinewentingwang/pytorch

Pull Requests Commits

feat(dockerfile): shrink layers & build cleaner (#95375)

georgettica committed 1 year ago

1c526664

Correct OneCycleLR doc example code to explicitly call optimizer.step() (#95730)

kiersten-stokes committed 1 year ago

60a1d295

[MPS] Add support for masked_scatter (#95743)

DenisVieriu97 committed 1 year ago

ed1957dc

[BE][DDPOptimizer] De-dup `p` and `param` (#95654)

awgu committed 1 year ago

d9cd9a13

Initial minifier smoke test + runbook (#95670)

mlazos committed 1 year ago

94bec94f

Remove dead ZeroGuard (#95701)

ezyang committed 1 year ago

7ea3aab4

[dynamo] Clear cache on dynamo dashboard accuracy tests (#95726)

williamwen42 committed 1 year ago

cf3638a9

Apply filter logic to disabled jobs dynamically (#95442)

huydhn committed 1 year ago

40d54cf8

[ONNX] Support 'dtype' argument for 'aten::norm' (#95637)

BowenBao committed 1 year ago

2fbbc336

hoist precomputed exprs from indices (#95690)

ngimel committed 1 year ago

88a31f4b

Warn on modification of OptimizedModule.forward (#95673)

wconstab committed 1 year ago

dc10ab15

Warn on dynamo OptimizedModule.forward() (#95672)

wconstab committed 1 year ago

6bdef7a5

Add support for Inductor + symbolic shapes + training (#93059)

ezyang committed 1 year ago

20dfce59

[jit] Add c++ stacktraces for jit::ErrorReport (#94842)

davidberard98 committed 1 year ago

70029214

Run tests in USE_PYTEST_LIST through run_tests (#95659)

clee2000 committed 1 year ago

e3c5c369

Rephrase zero_grad docs (#95643)

janeyx99 committed 1 year ago

e5b9d987

Build Triton in Docker image (#95233)

huydhn committed 1 year ago

ba43d908

Fix trymerge changed files count (#95720)

huydhn committed 1 year ago

b55d0d2a

Enable thp(transparent huge pages) for buffer sizes >=2MB (#93888)

snadampal committed 1 year ago

2cc845eb

[ao][fx] Enable observed -> quantized float for static quantized MultiheadAttention (#95636)

kev-zheng committed 1 year ago

f1dbfe2f

Clean up unused `fill_` sample inputs (#95117)

qqaatw committed 1 year ago

fafb4109

Add missing f-string specifiers (#95707)

ezyang committed 1 year ago

835122c8

Add standalone torch._inductor.compile() API (#95594)

jansel committed 1 year ago

e13b8041

[quant][pt2e] Add support for dynamic quantization with symmetric quant for input (#94854)

JacobSzwejbka committed 1 year ago

fc324d34

[dynamo] avoid truncation of python pointers (#95619)

cyyever committed 1 year ago

f8ad64d5

[dtensor][BE] remove redundant tests (#94838)

wanchaol committed 1 year ago

1e15a272

[dtensor] support creating DTensor in submesh (#95458)

wanchaol committed 1 year ago

2a1cb964

[dtensor] refactor get_coordiniate (#95457)

wanchaol committed 1 year ago

261eb46d

[dtensor] use tracing for metadata prop (#95456)

wanchaol committed 1 year ago

bb9a05b1

Enabling FlashAttention for SDPA when given NestedTensor (#95438)

drisspg committed 1 year ago

80614783

Newer Older