doc review compression-overview part

[KeLiChloe]

Description

Review on model compression 【overview】 part.

• 语法部分：

1. can be difficult to be deployed on ==》 which can be difficult to be deployed on
2. find the best compressed model ==》 find the best-compressed model
3. make it become smaller ==》makes it smaller
4. The overall compression pipeline in NNI. ==》 The overall compression pipeline in NNI is shown above.
5. The interface and APIs are unified for both PyTorch and TensorFlow, currently only PyTorch version has been supported, TensorFlow version will be supported in future. ==> The interface and APIs are unified for both PyTorch and TensorFlow. Currently only PyTorch version has been supported, and TensorFlow version will be supported in future.

• General Tips

1. Deep neural networks (DNNs) have achieved great success in many tasks，此处的tasks可以举一些例子，==》 in many tasks like embedded development and scenarios that needs rapid feedbacks.
2. overview的第一段简单说了两种模型压缩的方法，在段落末尾加上：We further elaborate on the two methods, pruning and quantization, in the following chapters 来做总结，同时引导读者继续阅读。另外，这里加上了图示来解释pruning和quantization的区别，一目了然。
3. Since NNI compression algorithms are not meant to compress model while NNI speedup tool can truly compress model and reduce latency.
   ==》Note that NNI pruners or quantizers are not meant to physically compact the model but for simulating the compression effect.
4. The algorithms include ==> The supported model compression algorithms include 指代更清晰
5. 在speed up章节加上了pipeline图示
6. 加上 If users want to apply both, sequential mode is recommended as common practice.

Checklist

• test case
• redundant weights or channels of layers 这里的 channels of layers指的是channel的什么？number吗？
• Users could further use NNI’s auto-tuning power to find the best compressed model, which is detailed in Auto Model Compression. 这里的Auto Model Compression是什么没说清楚，是NNI 的函数吗？还是doc的某个章节？

[ghost]

All CLA requirements met.