PulseAugur
EN
LIVE 21:01:46
ENTITY Logarithmic Learning Rate Decay

Logarithmic Learning Rate Decay

PulseAugur coverage of Logarithmic Learning Rate Decay — every cluster mentioning Logarithmic Learning Rate Decay across labs, papers, and developer communities, ranked by signal.

Show in brief
Total · 30d
0
1 over 90d
Releases · 30d
0
0 over 90d
Papers · 30d
0
1 over 90d
TIER MIX · 90D
TOPICS
RECENT · PAGE 1/1 · 1 TOTAL
  1. TOOL · CL_180649 ·

    New nGPT training method halves token requirements for MoE models

    Researchers have developed a new training methodology called normalized Transformer (nGPT) that constrains model parameters and activations to a unit hypersphere for improved representation learning. This recipe, detail…