RNN Record

提示：文章写完后，目录可以自动生成，如何生成可参考右边的帮助文档

---

前言

Today, I learnt Recurrent Neural Networks and Video Models.
Image captioning: input is a image, output is a sentence describing the image. Video classification is a many to one RNN, input is a sequence of images and output is a label.
Machine Translation is a many to many RNN. Input is a sequence of words, output is a sequence of words.

Per-frame video classificaition: sequence of images to sequence of labels.

---

提示：以下是本篇文章正文内容，下面案例可供参考

一、pandas是什么？

示例：pandas 是基于NumPy 的一种工具，该工具是为了解决数据分析任务而创建的。

二、使用步骤

1.Tanh

2.Sequence to Sequence

(Many to one) + (One to many)
Many to one: Encode input sequence in a single vector
One to many: Produce output sequence from single input vector.

---

总结

提示：这里对文章进行总结：
例如：以上就是今天要讲的内容，本文仅仅简单介绍了pandas的使用，而pandas提供了大量能使我们快速便捷地处理数据的函数和方法。