PAIR-code / tiny-transformers

Apache License 2.0
14 stars 2 forks source link

Provide a visualization for the norms of different parts of a transformer model #25

Open iislucas opened 1 month ago

iislucas commented 1 month ago

A common debugging method is to monitor the norm of some part of a ML model's weight and activation space.

Goal:

Code pointers: