Training Acceleration Trick Deployable via Weight Fusion

torchcompiled · x · 2026-07-15

This repost/quote discusses an interesting training trick: parameterizing weights as a "learned weighted blend of symexp and regular linear weights," which reportedly yields a 1.42x wall-clock training speedup.

The author also notes:

Related event: SEL Weight Reparameterization: ~1.42x Training Speedup, Folds Back to Standard Weights(7 posts)→

Original post →

More from Infra

Infra channel →