pytorch/benchmark

Pull Requests Commits

address feedback

Krovatkin committed 4 years ago

b73851e7

Krovatkin committed 4 years ago

936a20f9

check environment

Krovatkin committed 4 years ago

0e639a53

Krovatkin committed 4 years ago

9ca69a09

Add the access token to upload data to Scuba. (#236)

xuzhao9 committed 4 years ago

Verified b81825e5

Migrate CI pipelines to non-docker settings. (#233)

xuzhao9 committed 4 years ago

Verified 0340596f

fix version for now (#227)

Krovatkin committed 4 years ago

Verified a4ffcd68

Fix upload-scribe.sh path error, and migrate sweep pipeline to metal (#221)

xuzhao9 committed 4 years ago

Verified b0a18211

update unidash link

wconstab committed 4 years ago

Verified cac3b7f3

Fix a bug in CI script of computing score. (#215)

xuzhao9 committed 4 years ago

Verified 6f1e1b9b

Wconstab/install quiet (#202)

wconstab committed 4 years ago

Verified 40799b47

Update config v0 to metal node. (#205)

xuzhao9 committed 4 years ago

d7f12c91

Fix attention... name in score files

wconstab committed 4 years ago

0e4c9ade

Remove shortened names

wconstab committed 4 years ago

e2ec6ae2

Add dump_bench.py script for dumping json files (#207)

wconstab committed 4 years ago

Verified 60a751a6

Create get_torch_nightly_wheels.py (#182)

wconstab committed 4 years ago

Verified bb9a1b46

Clean up UI for pytest machine config (#204)

wconstab committed 4 years ago

Verified ccea180a

Clean up run_bench script (#197)

wconstab committed 4 years ago

Verified 2dc08cf7

Add machine_util.py for tuning optimal benchmark settings (#183)

wconstab committed 4 years ago

Verified f9e5a75a

Update the GH Actions file to migrate the CI to metal node. (#200)

xuzhao9 committed 4 years ago

Verified a56c9e61

Fix master CI by removing maskrcnn_benchmark and use pip instead of conda nightlies (#196)

wconstab committed 4 years ago

Verified c0c4d9dc

Add pip instructions to README (#195)

wconstab committed 4 years ago

Verified 8b8779db

Fix CI schedule time to benchmark the latest nightly. (#192)

xuzhao9 committed 4 years ago

Verified 0c34e941

Normalize demucs/maml behavior when using get_module (#190)

jansel committed 4 years ago

Verified 6f79061b

Make benchmarks run quietly for inference (#188)

jansel committed 4 years ago

Verified 8da374f7

Specifying eval mode in mobilenetv3 (#189)

navahgar committed 4 years ago

Verified 17922cdd

Fix a bug in pytorch_struct that raises NotImplemented error in the latest pytorch-nightly 20210107

wconstab committed 4 years ago

Verified 9cd2d3a1

Switch BERT, pytorch_struct, and fastNLP to develop mode (#187)

jansel committed 4 years ago

Verified f34fed66

Misc requirements and debuggability fixes (#185)

jansel committed 4 years ago

Verified a8c72f08

Enable 0.1 benchmarking build workflow. (#176)

xuzhao9 committed 5 years ago

Verified 5008a33d

Older