warner-benjamin/optimi

Issues

• Casting Existing FP32/FP16 model weights to BF16 + Kahan Summation

  #7 opened 2 months ago by zaptrem
  0

• [FeatureRequest]gradient-descent-the-ultimate-optimizer

  #6 opened 3 months ago by sdbds
  0

• When using accelerator with Gradient Release hook, Increasing VRAM consumption

  #5 opened 3 months ago by sdbds
  2

• Document error

  #4 opened 3 months ago by sdbds
  1

• Version 0.1.0 is gone from PyPi

  #1 opened 10 months ago by mensfeld
  2