Skip to content

MMCV Release V1.3.2
Compare
Choose a tag to compare
@zhouzaida zhouzaida released this 30 Apr 07:56
v1.3.2
ab973df
This commit was created on GitHub.com and signed with GitHub’s verified signature. The key has expired.
GPG key ID: 4AEE18F83AFDEB23
Expired
Learn about vigilant mode.

New Features

  • Add Rotated ROI align op for pytorch and onnxruntime (cpu) (#933)
  • Support Multi-Scale-DeformAttention in deformable-detr (#878)
  • Support CPU training with MMDataParallel (#972)
  • Support auto_fp16 using torch.cuda.amp when PyTorch >= 1.6.0 (#951)
  • Add PyTorch Profiler to analyze training process (#937)

Improvements

  • Remove redundant self.init_cfg (#959)
  • Refactor init cfg (#958)
  • Refactor readme.md (#984)
  • Use Sequential rather than nn.Sequential in build_model_from_cfg (#992)
  • Use len(result) as batch size in engine/test.py (#987)
  • Support clipping min_lr in StepLrUpdaterHook #995

Bug Fixes

  • Fix compile without cuda (#945)
  • Rename init_weight to init_weights (#971)
  • Incorrect citation in transformer documentation (#974)
  • Fix data type in fused-bias-leakyrelu for apex fp16 training (#981)
  • Fix raise error bug in registering multiple names (#949)
Assets 2
Loading