pytorch/benchmark

Pull Requests Commits

Default use throughput mode with cpu userbenchmark launcher

chuanqi129 committed 2 years ago

fb22910a

Enable throughputs metric in cpu userbenchmark

chuanqi129 committed 2 years ago

51a953dd

Add throughputs metric in torchbench model metrics

chuanqi129 committed 2 years ago

e1339903

Move torchrec_dlrm to canary models (#1725)

xuzhao9 committed 2 years ago

a933ccd9

Fix the failed unit test on A100 (#1720)

xuzhao9 committed 2 years ago

cb91cd1f

Migrate to use run.py

xuzhao9 committed 2 years ago

8526eabb

Moco Inplace Update Params (#1693)

eellison committed 2 years ago

66050c23

Add bisection for userbenchmarks (#1638)

xuzhao9 committed 2 years ago

db37ddda

Add cpu_peak_mem and gpu_peak_mem to the metrics

xuzhao9 committed 2 years ago

59dbaaaa

Improve the regression detector to generate bisection config (#1710)

xuzhao9 committed 2 years ago

1734a5d8

always upload stats to s3+scribe (#1711)

janeyx99 committed 2 years ago

db3e43ec

Migrate to use the correctness checking code in dynamo torchbench.py (#1700)

xuzhao9 committed 2 years ago

9a58c643

Add metrics param for cpu userbenchmark (#1674)

chuanqi129 committed 2 years ago

df27114e

Fix a bug in S3 upload script (#1684)

xuzhao9 committed 2 years ago

611bb442

Set the correct grad mode for train and eval correctness tests

xuzhao9 committed 2 years ago

c1408a3a

add sam to dense_models, retitle issues (#1688)

janeyx99 committed 2 years ago

0e663c85

Support downloading yaml files from S3 (#1685)

xuzhao9 committed 2 years ago

c596059b

Remove redundant dependency (#1686)

spzala committed 2 years ago

d40fff39

Set threshold to 30% and disable torch-nightly unstable models (#1681)

janeyx99 committed 2 years ago

3f2a2a15

Upload the regression file to Amazon S3 (#1679)

xuzhao9 committed 2 years ago

9504651c

Add feature to run nightly with regression yaml (#1677)

xuzhao9 committed 2 years ago

d0a05378

Add SAM (#1661)

msaroufim committed 2 years ago

cc53ec65

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

Older