ClippedScion

ClippedScion

Reference implementation of ClippedScion optimizers supporting multiple norms.

Reference implementation of ClippedScion and UnconstrainedClippedScion optimizers extending Scion with gradient clipping. Supports multiple norms and hyperparameter configurations for nanoGPT, CNN, and DeiT training with example setups and citation information.

Deep Neural NetworksOptimization
Key facts
Maturity
Support
C4DT
Inactive
Lab
Active
  • Technical

Laboratory for Information and Inference Systems

Laboratory for Information and Inference Systems
Volkan Cevher

Prof. Volkan Cevher

At LIONS, we are concerned with optimized information extraction from signals or data volumes. We therefore develop mathematical theory and computational methods for information recovery from highly incomplete data.

This page was last edited on 2026-03-03.