6.s191 on 111qqz的小窝https://111qqz.com/tags/6.s191/Recent content in 6.s191 on 111qqz的小窝Hugo -- gohugo.iozh-cnhust.111qqz@gmail.com (111qqz)hust.111qqz@gmail.com (111qqz)© 2011-2026 111qqzSun, 06 Sep 2026 22:00:00 +0800从 Forward 到 Optimizer：重建 Neural Network Training 的 Mental Modelhttps://111qqz.com/2026/09/nn-training-loop-mental-model/Sun, 06 Sep 2026 22:00:00 +0800hust.111qqz@gmail.com (111qqz)https://111qqz.com/2026/09/nn-training-loop-mental-model/<h2 class="relative group">起因 <div id="起因" class="anchor"></div> <span class="absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none"> <a class="text-primary-300 dark:text-neutral-700 !no-underline" href="#%e8%b5%b7%e5%9b%a0" aria-label="锚点">#</a> </span> </h2> <p>最近在继续 CS336 的过程中，发现一个问题：很多模型训练侧的基础概念，虽然&quot;知道&quot;，但已经不像以前那么顺手了。</p>