Then you have a decoder, which roughly speaking is the encoder in reverse.

然后你有一个解，它大致上是编的反向操作。

So that's called the encoder stage.

就是编阶段。

Let's first explore the optical rotary encoder.

让我们首先来探索一下光学旋转编。

By the way, this is technically called an incremental quadrature encoder.

顺便说一下，在技术上被称为增量式正交编。

The encoder of the transformer is going to process all of these tokens simultaneously to predict the best character and the best word eventually.

Transformer 模型的编将同时处理所有解析单元，以最终预测最佳字符和最佳单词。

Scroll wheels commonly use either optical rotary encoders or magnetic rotary position sensors

滚轮通常使用光学旋转编或磁性旋转位置传感。

None of the vector-to-text stuff works unless you have this assumption of knowing the encoder and also being able to make a lot of queries to it.

除非你知道编，并且能够对其提出大量查询，否则所有将向量转换为文本的方法都不会奏效。

The key to the optical rotary encoder is that, as the wheel is rotated, the light emitted from the IR LED is intermittently blocked by the encoding disk.

光学旋转编的关键在于，当滚轮旋转时， 从红外LED发出的光线会被编盘间歇性地遮挡。

One thing to note is that in this video we cover just two methods to measure rotational motion, an optical encoder and a magnetic position encoder.

需要注意的是， 在段视频中我们仅介绍了两种测量旋转运动的方法，一种是光学编， 另一种是磁性位置编。

However, it is possible to have an absolute optical encoder, which has a disk that looks like this, where different angles have different sections open to let the light through.

然而， 也有可能使用绝对式光学编，它有一个看起来像样的盘片，不同的角度有不同的区域可以让光线通过。

First, the incremental optical encoder from earlier can't read out an absolute angular position because the wheel has 48 equally spaced spokes, and it can't discern one angle from the next.

首先， 之前的增量式光学编无法读取绝对角度位置， 因为滚轮有48个等间距的辐条，它无法区分一个角度和下一个角度。

You input an A, and it might output a C. This may not sound like much, but you must know that the most powerful part of the Enigma is the core encryption part, the encoder's core is a gear.

您输入 A， 它可能会输出 C。 听起来可能不多，但你必须知道 Enigma 最强大的部分是核心加密部分，编的核心是一个齿轮。