前言
我們來做 How Google does Machine Learning 的第二個lab~
這次鐵人賽的30天中，我目前所寫文章的所有課程目錄如下：
- 【Day 1】準備日 – 註冊coursera與訂閱課程
- Course – How Google does Machine Learning
- Chapter 1 – Introduction to specialization
- Chapter 2 – What it means to be AI first
- Chapter 3 – How Google does ML
- Chapter 5 – Python notebooks in the cloud
Course – How Google does Machine Learning
第五章節的課程地圖：(紅字標記為本篇文章中會介紹到的章節)
* Python notebooks in the cloud
* Module Introduction
* Cloud Datalab
* Cloud Datalab
* Demo: Cloud Datalab
* Development process
* Demo of rehosting Cloud Datalab
* Working with managed services
* Computation and storage
* Lab: Rent-a-VM
* Intro to Qwiklabs
* Intro to Renting-VM Lab
* Lab: Rent-a-VM to process earthquake data
* Lab debrief
* Cloud Shell
* Third wave of cloud
* Third Wave of Cloud: Fully-Managed Services
* Third Wave of Cloud: Serverless Data Analysis
* Third Wave of Cloud: BigQuery and Cloud Datalab
* Datalab and BigQuery
* Lab Intro: Analyzing data using Datalab and BigQuery
* Lab: Analyzing data using Datalab and BigQuery
* Lab Debrief: Analyzing Data using Datalab and BigQuery
* Machine Learning with Sara Robinson
* ML, not rules
* Pre-trained ML APIs
* Vision API in action
* Video intelligence API
* Cloud Speech API
* Translation and NL
* Lab: Machine Learning APIs
* Lab: Pretrained ML APIs Intro
* Lab: Invoking Machine Learning APIs
* Lab Solution
Lab: Machine Learning APIs
課程地圖
* Python notebooks in the cloud
* Lab: Machine Learning APIs
* Lab: Pretrained ML APIs Intro
* Lab: Invoking Machine Learning APIs
* Lab Solution
在這個lab中，我們將實作之前 【Day 12】- google圖片辨識(Vision), 影片辨識(Video), 語音辨識, 語言翻譯, 自然語言辨識(NL) API功能總整理 的API應用
我們會示範如何進行API串接，讓我們能直接使用google已訓練好的ML API來直接實現功能。
part 0 : (事前準備) 開啟 GCP console
- 請先參考 【Day 9】- 每次在google雲端上開始lab前都要的事前準備與注意事項 的內容，完成到運行中階段。
part 1 : (建立機器) 建立並執行 Cloud Datalab VM
Step 0 : 打開 Cloud Shell
- 如果不清楚 Cloud Shell 如何開啟，請參考 【Day 11】- Cloud Shell 的介紹與 google雲的三代變化, 使用ML與一般演算法的比較與優勢
Step 1 : 首先我們要先知道我們的 compute zones 在哪，我們可以透過以下指令知