pytorch/benchmark

Pull Requests Commits

Disble non-functional v2 workflow

ZainRizvi committed 216 days ago

3a5fb937

Disable v3-nightly.yml

ZainRizvi committed 216 days ago

Verified 6141dd96

Simplify disabling of the helper functions on tensor properties (#155259)

anijain2305 committed 241 days ago

f144fe90

Revert "Record the pre-graph bytecode using fast record function event (#154974)"

generatedunixname499836121 committed 243 days ago

68cf9dff

Record the pre-graph bytecode using fast record function event

anijain2305 committed 244 days ago

6bb98539

Mark a vt unspecialized nn module variable source earlier

anijain2305 committed 244 days ago

73c93cf7

Incorporate coalesce analysis in codegen (#153751)

eellison committed 245 days ago

eee7b1fc

Vary batch size when running dynamic shapes benchmarks (#154805)

bobrenjc93 committed 246 days ago

961a7682

delete code that died a long time ago (#154802)

bobrenjc93 committed 247 days ago

a1de07bc

Fix typo in dynamo utils #154639 (#154748)

generatedunixname499836121 committed 247 days ago

7efb806d

Improve dynamo utils typing with TypeIs and TypeGuard (#154639)

generatedunixname499836121 committed 247 days ago

2b9e996b

Trace into descriptor with `__set__` (#154176)

StrongerXi committed 249 days ago

6dd42b53

Fix the benchmark CI failure due to Opacus 1.5.4 (#2621)

HuanyuZhang committed 251 days ago

9486136d

Pin opacus to 1.5.3 on opacus_cifar10 model (#2620)

atalman committed 252 days ago

e03a63be

update torchvision pin (#154255)

BoyuanFeng committed 252 days ago

2caf997d

Use cuBLAS default workspace size in Lt (#153556)

generatedunixname499836121 committed 253 days ago

e759c474

Fix CHROMIUM_EVENT_LOG being none

jamesjwu committed 256 days ago

4617d91c

Don't upload compiler benchmark debug info to the benchmark database (#153769)

huydhn committed 256 days ago

6693f584

fix anti-pattern for cudagraph (#2619)

BoyuanFeng committed 257 days ago

227fb01e

Allow higher fp16 tolerance for phlippe_resnet on CUDA 12.8 (#154109)

huydhn committed 257 days ago

1b4c3199

Extend empty_gpu_cache to mps (#153657)

malfet committed 264 days ago

316981ee

Fix test broken by D73809989 (#153413)

masnesral committed 265 days ago

e32ebfe2

Optimize dedupe region ancestor tracking (#152589)

mlazos committed 265 days ago

f3ef83db

Fix misleadingly high AOT Inductor dashboard performance (#153060)

generatedunixname499836121 committed 265 days ago

d36d4778

Use universal flatten APIs (#152505)

mlazos committed 265 days ago

ecf479d0

Track node mutations (#152389)

mlazos committed 265 days ago

7345ba37

Revert "Track node mutations (#152389)"

generatedunixname499836121 committed 266 days ago

99082650

Revert "Use universal flatten APIs (#152505)"

generatedunixname499836121 committed 266 days ago

595677cd

Update ruff to 0.11.8 (#153249)

generatedunixname499836121 committed 267 days ago

9e0e6b2a

Use universal flatten APIs (#152505)

mlazos committed 268 days ago

ee027de2

Older