Last week we released NanoGPT Slowrun , an open repo for data-efficient learning algorithms. The rules are simple: train on 100M tokens from FineWeb, use as much compute as you want, lowest validation loss wins. Improvements are submitted as PRs to the repo and merged if they lower val loss. The constraint is the inverse of speedruns like modded-nanogpt , which optimize wall-clock time. Those benchmarks have been hugely productive, but optimizing for speed filters out expensive ideas: heavy regularization, second-order optimizers, gradient descent alternatives. Slowrun is built for exactly those ideas.
:first-child]:h-full [&:first-child]:w-full [&:first-child]:mb-0 [&:first-child]:rounded-[inherit] h-full w-full。业内人士推荐PDF资料作为进阶阅读
。关于这个话题，PDF资料提供了深入分析
ВсеПолитикаОбществоПроисшествияКонфликтыПреступность
MiniMax 同样在证明自己的造血能力。2 月 13 日发布的 MiniMax M2.5 文本模型，上线不到一周便迅速登顶 OpenRouter 的 Token 调用榜首。在 2 月 9 日至 15 日的统计周期内，仅 M2.5 一个模型就贡献了 1.44T 的 Token 调用量，规模超过了 Kimi K2.5、GLM-5 和 DeepSeek V3.2 的总和。。关于这个话题，PDF资料提供了深入分析