Code

Update fp16 models and performance#178

Code

Merged

ZwwWayne merged 9 commits intoopen-mmlab:masteropen-mmlab/mmdetection3d:masterfrom ZwwWayne:support-fp16ZwwWayne/mmdetection3d:support-fp16Copy head branch name to clipboardOct 30, 2020

+15-10

Commits

Commits on Sep 24, 2020

• Support to train using FP16

  ZwwWayne

  committed
  506b11a

  Copy full SHA for 506b11a

Commits on Sep 25, 2020

• fix type inconsistency error on naive syncBN

  ZwwWayne

  committed
  b76fb3f

  Copy full SHA for b76fb3f

Commits on Oct 1, 2020

• resolve comments

  ZwwWayne

  committed
  44826f6

  Copy full SHA for 44826f6

Commits on Oct 7, 2020

• Merge branch 'master' of github.com:open-mmlab/mmdetection3d into support-fp16

  ZwwWayne

  committed
  114cd8f

  Copy full SHA for 114cd8f

Commits on Oct 10, 2020

• clean nan check

  ZwwWayne

  committed
  a2ae673

  Copy full SHA for a2ae673

Commits on Oct 23, 2020

• clean nan check

  ZwwWayne

  committed
  cbac830

  Copy full SHA for cbac830

• Merge branch 'master' of github.com:open-mmlab/mmdetection3d into support-fp16

  Show description for a2cbb9f
  ZwwWayne

  committed
  a2cbb9f

  Copy full SHA for a2cbb9f

• clean unnecessary modification

  ZwwWayne

  committed
  6a18a68

  Copy full SHA for 6a18a68

Commits on Oct 30, 2020

• update model link

  ZwwWayne

  committed
  bf9fdaa

  Copy full SHA for bf9fdaa