pytorch/benchmark

Pull Requests Commits

Add metrics param for cpu userbenchmark

chuanqi129 committed 2 years ago

09745b16

Detect both speedup and regressions in v3 workflow (#1667)

xuzhao9 committed 2 years ago

08d2fd5c

Fix batch size for vision_maskrcnn (#1656)

ezyang committed 2 years ago

6ddb3d05

Use CUDA 11.8 versions as the default (#1652)

xuzhao9 committed 2 years ago

987bbec3

Update README.md (#1660)

H-Huang committed 2 years ago

d5fd2d14

Fix TypeError in the e2e hf_t5 (#1655)

spzala committed 2 years ago

31dc5dfe

Reduce nanogpt gen sizes (#1659)

drisspg committed 2 years ago

f4acd1a7

Fix a bug in cpu userbenchmark with launcher (#1650)

chuanqi129 committed 2 years ago

3691415e

Sync the timm repo revision with the upstream pytorch/pytorch repo (#1651)

xuzhao9 committed 2 years ago

44ccad1d

Remove conda env activation when tuning gpu (#1646)

xuzhao9 committed 2 years ago

6c4b5de4

Remove local keyword from build script (#1642)

xuzhao9 committed 2 years ago

ec359fad

More dense models from the newly added ones (#1643)

janeyx99 committed 2 years ago

0349e8b3

Enhance cpu userbenchmark deal with failures and timeout (#1635)

chuanqi129 committed 2 years ago

b3328994

Increase delta, exclude CPU reporting (#1641)

janeyx99 committed 2 years ago

7a53ab4b

add newly models to dense list to not run on sparseadam (#1639)

janeyx99 committed 2 years ago

34fc32f6

use g++ 10 for torchrec_dlrm (#1622)

xuzhao9 committed 2 years ago

88a40a73

Regression detector should print correct repr (#1640)

janeyx99 committed 2 years ago

31660bbb

Add nanogpt model (#1628)

drisspg committed 2 years ago

a9469073

Update vocab size for LLAMA (#1634)

msaroufim committed 2 years ago

06e47a91

Add workflow on AWS C5 24xlarge (#1637)

xuzhao9 committed 2 years ago

d091a9dc

Add some better GNN benchmarks from pytorch_geometric (#1633)

ezyang committed 2 years ago

a98fa65c

Add filter example to the pytest-benchmark driver section (#1632)

spzala committed 2 years ago

3480ee06

Dont fail on training with random weights (#1631)

anijain2305 committed 2 years ago

ea0bdfe6

Add hf_GPT2_generate and hf_T5_generate (#1625)

ezyang committed 2 years ago

18ceb5d7

Delay the nightly and daily workflows by 2 hours (#1627)

xuzhao9 committed 2 years ago

2788e24e

Add cm3leon (#1620)

ezyang committed 2 years ago

7d467c0a

Fix pytorch branch ref to run userbenchmark instruction-count (#1623)

spzala committed 2 years ago

0f28344b

...detect speedups too (#1621)

janeyx99 committed 2 years ago

33128f16

Remove downloading the nightly, check only if the workflow is scheduled (#1613)

xuzhao9 committed 2 years ago

fd53b084

add delta to denoise in regression detector (#1616)

janeyx99 committed 2 years ago

065c3580

Older