Nonetheless worry not, we’re going to use an API from GCP known as App Engine. It is a completely managed serverless platform by Google. This means you solely have to stress in regards to the code, not the infrastructure. App Engine helps (Node.js, Java, Ruby, C#, Go, Python, or PHP) and it scales up and down upon demand and in addition you pay for what you make the most of. I KNOW!
Okay, for the sake of this practice, we’re going to put collectively our information using the straight-line equation Y = Mx + C. X could be the enter to the model and Y could be the output. We’ll create 5000 information components in a single file and save them as .npy recordsdata in a separate folder. I’ll have carried out this inside the teaching file as successfully nevertheless I would love you to know how one can add and cargo completely totally different recordsdata contained in the App Engine ambiance.
import numpy as np
import os# Parameters for the straight line equation
m = 2.5 # Slope
c = 1.0 # Intercept
# Generate x values
x = np.linspace(0, 1000, 5000)
print(len(x))
# Generate y values using the straight line equation
y = m * x + c
# Save the data to .npy recordsdata
np.save(os.path.be part of(os.path.dirname(__file__),'Information','x.npy'),x)
np.save(os.path.be part of(os.path.dirname(__file__),'Information','y.npy'),y)
Throughout the subsequent file, you may load this data and put together a TensorFlow model (Pytorch fashions might be hosted equally) that will predict the output amount given an enter amount and might try and mimic the habits of a straight-line equation.
I do know,
- We’re to not make use of ML for duties whose exact choices are recognized nevertheless that’s merely to take care of the model simple.
- To utilize ML for this exercise is like using a Jackhammer to crack a nut nevertheless please…
Throughout the teaching file, you may load the data, define a model, compile it, after which put it apart after teaching. Proper right here is the code to do this,
import tensorflow as tf
from tensorflow.keras.fashions import Sequential
from tensorflow.keras.layers import Dense
import numpy as np
import os# Load the data
x = np.load(os.path.be part of(os.path.dirname(__file__),'Information','x.npy'))
y = np.load(os.path.be part of(os.path.dirname(__file__),'Information','y.npy'))
# Define the neural group model
model = Sequential([
Dense(2, activation='relu', input_shape=(1,)),
Dense(4, activation='relu'),
Dense(1)
])
# Compile the model
model.compile(optimizer='adam', loss='mse')
# Put together the model
model.match(x, y, epochs=10)
# make a sample request to see the desired outcomes are being achieved
req = 10
print('model prediction:',model.predict(np.array([[float(req)]])))
print('Precise reply:',2.5*req+1.0)
# Save the model
model.save(os.path.be part of(os.path.dirname(__file__),'linear_model.h5'))
After working the above code, the following outcomes are obtained.
In a model new file, we’re going to load the saved model, make predictions, and description the URLs for serving by the use of our app. We’ll use Flask for this software program (a framework to assemble and deploy Python apps). A URL will likely be set inside the Flask app, which we’re going to hit using CURL, and the model’s prediction shall be obtained inside the kind of a JSON object. Make certain that to acquire Flask using pip or conda, nonetheless, you make the most of and deal with your environments.
It is important to establish this file app.py to work with Flask.
from flask import Flask, request, jsonify
import tensorflow as tf
import numpy as np
import osapp = Flask(__name__)
# Load the expert model
model = tf.keras.fashions.load_model(os.path.be part of(os.path.dirname(__file__),
'linear_model.h5'))
@app.route('/predict', methods=['GET'])
def predict():
try:
# Get the x price from the request
x_value = float(request.args.get('x'))
# Reshape the enter for the model
x_input = np.array([[x_value]])
# Predict using the model
prediction = model.predict(x_input)
# Return the prediction as a JSON response
return jsonify({'y_intercept': float(prediction[0][0])})
apart from Exception as e:
return jsonify({'error': str(e)})
if __name__ == '__main__':
app.run(debug=True)
