In this lesson, you will train an ML model on a data set that's out of this world: _UFO sightings over the past century_, sourced from [NUFORC's database](https://www.nuforc.org).
We will continue our use of notebooks to clean data and train our model, but you can take the process one step further by exploring using a model 'in the wild', so to speak: in a web app.
There are several ways to build web apps to consume machine learning models. Your web architecture may influence the way your model is trained. Imagine that you are working in a business where the data science group has trained a model that they want you to use in an app.
- **Is it a web app or a mobile app?** If you are building a mobile app or need to use the model in an IoT context, you could use [TensorFlow Lite](https://www.tensorflow.org/lite/) and use the model in an Android or iOS app.
- **Where will the model reside**? In the cloud or locally?
- **Offline support**. Does the app have to work offline?
- **What technology was used to train the model?** The chosen technology may influence the tooling you need to use.
- **Using Tensor flow**. If you are training a model using TensorFlow, for example, that ecosystem provides the ability to convert a TensorFlow model for use in a web app by using [TensorFlow.js](https://www.tensorflow.org/js/).
- **Using PyTorch**. If you are building a model using a library such as [PyTorch](https://pytorch.org/), you have the option to export it in [ONNX](https://onnx.ai/) (Open Neural Network Exchange) format for use in JavaScript web apps that can use the [Onnx Runtime](https://www.onnxruntime.ai/). This option will be explored in a future lesson for a Scikit-learn-trained model.
- **Using Lobe.ai or Azure Custom vision**. If you are using an ML SaaS (Software as a Service) system such as [Lobe.ai](https://lobe.ai/) or [Azure Custom Vision](https://azure.microsoft.com/services/cognitive-services/custom-vision-service/?WT.mc_id=academic-15963-cxa) to train a model, this type of software provides ways to export the model for many platforms, including building a bespoke API to be queried in the cloud by your online application.
You also have the opportunity to build an entire Flask web app that would be able to train the model itself in a web browser. This can also be done using TensorFlow.js in a JavaScript context.
For our purposes, since we have been working with Python-based notebooks, let's explore the steps you need to take to export a trained model from such a notebook to a format readable by a Python-built web app.
For this task, you need two tools: Flask and Pickle, both of which run on Python.
对于此任务，你需要两个工具：Flask和Pickle，它们都在Python上运行。
✅ What's [Flask](https://palletsprojects.com/p/flask/)? Defined as a 'micro-framework' by its creators, Flask provides the basic features of web frameworks using Python and a templating engine to build web pages. Take a look at [this Learn module](https://docs.microsoft.com/learn/modules/python-flask-build-ai-web-app?WT.mc_id=academic-15963-cxa) to practice building with Flask.
✅ What's [Pickle](https://docs.python.org/3/library/pickle.html)? Pickle 🥒 is a Python module that serializes and de-serializes a Python object structure. When you 'pickle' a model, you serialize or flatten its structure for use on the web. Be careful: pickle is not intrinsically secure, so be careful if prompted to 'un-pickle' a file. A pickled file has the suffix `.pkl`.
In this lesson you'll use data from 80,000 UFO sightings, gathered by [NUFORC](https://nuforc.org) (The National UFO Reporting Center). This data has some interesting descriptions of UFO sightings, for example:
- **Long example description**. "A man emerges from a beam of light that shines on a grassy field at night and he runs towards the Texas Instruments parking lot".
- **Short example description**. "the lights chased us".
The [ufos.csv](./data/ufos.csv) spreadsheet includes columns about the `city`, `state` and `country` where the sighting occurred, the object's `shape` and its `latitude` and `longitude`.
In the blank [notebook](notebook.ipynb) included in this lesson:
在包含在本课中的空白[notebook](notebook.ipynb)中：
1. import `pandas`, `matplotlib`, and `numpy` as you did in previous lessons and import the ufos spreadsheet. You can take a look at a sample data set:
@ -70,7 +70,7 @@ In the blank [notebook](notebook.ipynb) included in this lesson:
ufos.Country.unique()
```
1. Now, you can reduce the amount of data we need to deal with by dropping any null values and only importing sightings between 1-60 seconds:
3. 现在，你可以通过删除任何空值并仅导入1-60秒之间的目击数据来减少我们需要处理的数据量：
```python
ufos.dropna(inplace=True)
@ -80,9 +80,9 @@ In the blank [notebook](notebook.ipynb) included in this lesson:
ufos.info()
```
1. Import Scikit-learn's `LabelEncoder` library to convert the text values for countries to a number:
4. 导入Scikit-learn的`LabelEncoder`库，将国家的文本值转换为数字：
✅ LabelEncoder encodes data alphabetically
✅ LabelEncoder按字母顺序编码数据
```python
from sklearn.preprocessing import LabelEncoder
@ -92,7 +92,7 @@ In the blank [notebook](notebook.ipynb) included in this lesson:
ufos.head()
```
Your data should look like this:
你的数据应如下所示：
```output
Seconds Country Latitude Longitude
@ -103,11 +103,11 @@ In the blank [notebook](notebook.ipynb) included in this lesson:
24 3.0 3 51.783333 -0.783333
```
## Exercise - build your model
## 练习 - 建立你的模型
Now you can get ready to train a model by diving the data into the training and testing group.
现在，你可以通过将数据划分为训练和测试组来准备训练模型。
1. Select the three features you want to train on as your X vector, and the y vector will be the `Country`. You want to be able to input `Seconds`, `Latitude` and `Longitude` and get a country id to return.
The model you created isn't very revolutionary as you should be able to infer a `Country` from its `Latitude` and `Longitude`, but it's a good exercise to try to train from raw data that you cleaned, exported, and then use this model in a web app.
Now, it's time to _pickle_ your model! You can do that in a few lines of code. Once it's _pickled_, load your pickled model and test it against a sample data array containing values for seconds, latitude and longitude,
1. In that folder create three more folders: **static**, with a folder **css** inside it, and **templates`**. You should now have the following files and directories:
@ -170,9 +170,9 @@ Now you can build a Flask app to call your model and return similar results, but
ufo-model.pkl
```
✅ Refer to the solution folder for a view of the finished app
✅ 请参阅解决方案文件夹以查看已完成的应用程序
1. The first file to create in _web-app_ folder is **requirements.txt** file. Like _package.json_ in a JavaScript app, this file lists dependencies required by the app. In **requirements.txt** add the lines:
1. Now, you're ready to create three more files to finish the app:
6. 现在，你已准备好创建另外三个文件来完成应用程序：
1. Create **app.py** in the root
2. Create **index.html** in _templates_ directory.
3. Create **styles.css** in _static/css_ directory.
1. 在根目录中创建**app.py**
2. 在_templates_目录中创建**index.html**。
3. 在_static/css_目录中创建**styles.css**。
1. Build out the _styles.css__ file with a few styles:
7. 使用一些样式构建_styles.css_文件：
```css
body {
@ -233,7 +233,7 @@ Now you can build a Flask app to call your model and return similar results, but
}
```
1. Next, build out the _index.html_ file:
8. 接下来，构建_index.html_文件：
```html
<!DOCTYPE html>
@ -268,11 +268,11 @@ Now you can build a Flask app to call your model and return similar results, but
</html>
```
Take a look at the templating in this file. Notice the 'mustache' syntax around variables that will be provided by the app, like the prediction text: `{{}}`. There's also a form that posts a prediction to the `/predict` route.
Finally, you're ready to build the python file that drives the consumption of the model and the display of predictions:
最后，你已准备好构建使用模型和显示预测的python 文件：
1. In `app.py` add:
9. 在`app.py`中添加:
```python
import numpy as np
@ -309,39 +309,39 @@ Now you can build a Flask app to call your model and return similar results, but
app.run(debug=True)
```
> 💡 Tip: when you add [`debug=True`](https://www.askpython.com/python-modules/flask/flask-debug-mode) while running the web app using Flask, any changes you make to your application will be reflected immediately without the need to restart the server. Beware! Don't enable this mode in a production app.
If you run `python app.py` or `python3 app.py` - your web server starts up, locally, and you can fill out a short form to get an answer to your burning question about where UFOs have been sighted!
Before doing that, take a look at the parts of `app.py`:
在此之前，先看一下`app.py`的实现：
1. First, dependencies are loaded and the app starts.
1. Then, the model is imported.
1. Then, index.html is rendered on the home route.
1. 首先，加载依赖项并启动应用程序。
2. 然后，导入模型。
3. 然后，在home路由上渲染index.html。
On the `/predict` route, several things happen when the form is posted:
在`/predict`路由上，当表单被发布时会发生几件事情：
1. The form variables are gathered and converted to a numpy array. They are then sent to the model and a prediction is returned.
2. The Countries that we want displayed are re-rendered as readable text from their predicted country code, and that value is sent back to index.html to be rendered in the template.
Using a model this way, with Flask and a pickled model, is relatively straightforward. The hardest thing is to understand what shape the data is that must be sent to the model to get a prediction. That all depends on how the model was trained. This one has three data points to be input in order to get a prediction.
In a professional setting, you can see how good communication is necessary between the folks who train the model and those who consume it in a web or mobile app. In our case, it's only one person, you!
Instead of working in a notebook and importing the model to the Flask app, you could train the model right within the Flask app! Try converting your Python code in the notebook, perhaps after your data is cleaned, to train the model from within the app on a route called `train`. What are the pros and cons of pursuing this method?
There are many ways to build a web app to consume ML models. Make a list of the ways you could use JavaScript or Python to build a web app to leverage machine learning. Consider architecture: should the model stay in the app or live in the cloud? If the latter, how would you access it? Draw out an architectural model for an applied ML web solution.
feiyun0112
4 years ago
committed by
GitHub