EvoNorm

Unofficial PyTorch Implementation of EvoNorm

MIT License

Stars
122
Committers
2

Evolving Normalization-Activation Layers

Google AI and DeepMind

  • Implement EvoNorm S0 and B0 with Training Mode support
  • Solve Shape Error with group_std and instance_std functions
  • Solve NaN Error Issue with S0
  • Fix Error with shape in running variance calculation in EvoNorm B0
  • Solve NaN Error Issue with B0

Usage:

from evonorm2d import EvoNorm2D
# For B0 version
evoB0 = EvoNorm2D(input, affine = True, version = 'B0', training = True)

# For S0 version 
evoS0 = EvoNorm2D(input)