微软 MAI-Base-2 完成预训练，MAI-Base-3 训练周期已在筹备
MICROSOFT 🔥: MAI-Base-2 pre-training has been completed; MAI-Base-3 training cycle is already being prepared as well.
微软 MAI-Base-2 预训练已完成，MAI-Base-3 的训练周期也已在筹备中。据引用内容，MAI-Base-2 在留出评测上优于其测试过的所有开放权重模型，预训练 FLOPs 效率比最强的开放基础模型 DeepSeek-V4-Pro-Base 高 4-12 倍，微软将这部分节省用于训练更大的模型。Mustafa Suleyman 表示，MAI 已完全自研出 3T 参数级别的 SOTA 预训练基础模型，后训练结果即将公布。
查看原帖We now have state-of-the-art pre-trained base models in the 3T-parameter weight class, built entirely in-house at MAI. Post-training results coming soon -- stay tuned! Lots of details in our technical blog post. https://t.co/xcZ5uV4TAp
正文
当前显示原文，中文全文译文尚未就绪。
MICROSOFT 🔥: MAI-Base-2 pre-training has been completed; MAI-Base-3 training cycle is already being prepared as well.
> "MAI-Base-2 outperforms every open-weight model we tested on our held-out evaluations. It is 4-12x more pre-training FLOPs-efficient than the strongest open base model, DeepSeek-V4-Pro-Base (see Figure 3 below), and we have put that saving into training a larger model."
* Figure 1 shows the performance of MAI-Base-2 and other models with at least one trillion total parameters across a range of pre-training evaluations (lower is better).