NanoGPT Speedrun Record Falls to 67.6s With Canonical Token Masking (-15 Training Steps)

kellerjordan0 · x · 2026-09-19

Jan Varho set a new NanoGPT Speedrun world record of 67.6s (-0.4s) in modded-nanogpt with a simple idea: canonical token masking.

A tokenization-level observation translates into real training speedup — simple idea, clear payoff.

Original post →

More from Research

Research channel →