notes

Personal notes
git clone git://git.laack.co/notes.git
Log | Files | Refs

GradientClipping.md (789B)


      1 # Gradient Clipping
      2 
      3 ML P569
      4 
      5 **Definition:** Gradient clipping is the process of clipping gradients during backpropogration so they never exceed some threshold.
      6 
      7 This is another technique used to resolve issues relating to [ExplodingGradients](ExplodingGradients.md) particularly for RNNs where batch normalization does not work.
      8 
      9 There are two ways to do gradient clipping either with a threshold cut off or with vector scaling. With vector scaling we retain the direction of the vector and set the minimize the largest value to 1 (if greater than 1) while scaling all other features proprotionally. More commonly, we simply truncate values so if we have [100, .1] with a threshold of (-1,1) we would then scale the vector to [1, .1].
     10 
     11 Scaling the entire vector is called normalization.