Siraj Raval llSourcell

Senior AI/ML Engineer | Specializing in production-ready, trustworthy AI systems (GenAI, RAG, CV, MLOps).

llSourcell / tf.js

Created April 18, 2018 01:10

	<html>
	<head>
	<!-- Load TensorFlow.js -->
	<script src="https://cdn.jsdelivr.net/npm/@tensorflow/tfjs@0.9.0"> </script>

	<!-- Place your code in the script tag below. You can also use an external .js file -->
	<script>
	// Notice there is no 'import' statement. 'tf' is available on the index-page
	// because of the script tag above.

llSourcell / model.js

Created April 18, 2018 01:31

	// Define function
	function predict(input) {
	// y = a * x ^ 2 + b * x + c
	// More on tf.tidy in the next section
	return tf.tidy(() => {
	const x = tf.scalar(input);

	const ax2 = a.mul(x.square());
	const bx = b.mul(x);
	const y = ax2.add(bx).add(c);

llSourcell / code.js

Created April 18, 2018 01:32

	const model = tf.sequential();
	model.add(
	tf.layers.simpleRNN({
	units: 20,
	recurrentInitializer: 'GlorotNormal',
	inputShape: [80, 4]
	})
	);

	const optimizer = tf.train.sgd(LEARNING_RATE);

llSourcell / fraud_detection.py

Created April 24, 2018 22:58

	## Rule based approach
	## pseudocode

	def traditional_fraud_detection(user_purchase_transaction):
	purchase = user_purchase_transaction

	if (purchase.location == Zucks_office) & (purchase.amount > 1000)
	return false
	else
	return true

llSourcell / credit.py

Created April 24, 2018 23:38

	import pandas
	from sklearn.tree import DecisionTreeClassifier

	#Read transaction data
	data=pd.read_csv('../input/creditcard.csv')
	x_train, x_test, y_train, y_test = train_test_split(data, target, train_size = 0.50)

	#train model
	model = DecisionTreeClassifier.fit(x_train,y_train)
	score = model.score(x_test, y_test)

llSourcell / regression.py

Created April 24, 2018 23:39

	import pandas
	from sklearn.tree import LogisticRegressionClassifier

	#Read transaction data
	data=pd.read_csv('../input/creditcard.csv')
	x_train, x_test, y_train, y_test = train_test_split(data, target, train_size = 0.50)

	#train model
	model = LogisticRegressionClassifier.fit(x_train,y_train)
	score = model.score(x_test, y_test)

llSourcell / Cover_letter_example

Created June 27, 2018 01:32

	Hello Stack Overflow,

	I’ve used Stack Overflow for as long as I’ve been a developer,
	and I recently came across a post about the architecture of your
	products on Nick Craver’s blog. It made me think,
	“I really want to work with these people who care so much about what they do.”
	I’m super excited to hear about all the tools you have built to make developer
	processes more streamlined; that’s right up my alley.

	At my current job I started out as a web dev, b

llSourcell / descent.py

Last active July 4, 2018 01:18


	//Our linear regression model is y = mx + b
	//Our parameters are thus m and b. What are the optimal values...
	//Lets use gradient descent to find out!


	def linear_regression(X, y, m_current=0, b_current=0, epochs=1000, learning_rate=0.0001):
	N = float(len(y))
	for i in range(epochs):
	y_current = (m_current * X) + b_current

llSourcell / style_transfer.py

Created July 4, 2018 22:38

	from keras.applications import vgg19
	from keras import backend as K
	from scipy.optimize import fmin_l_bfgs_b

	# Define the base image and style image
	base_image_path = base_image_path
	style_reference_image_path = style_reference_image_path

	# these are the weights of the different loss components
	total_variation_weight = 1.0