Day #120 - Tensorflow

Tensor (N-Dimensional array of data)

• High Performance Library for Numerical Computation
• Represented as Directed Graphs
• DAG (Directed Acyclic Graphs)
• Edges - Arrays of Data

DAG

• Language independent version of representation
• Similar to JVM
• Tensorflow engine written in C++
• Tensorflow Lite - On device interference of ML Models

API Hierarchy

• Number of abstraction layers
• High Level API -> tf.estimator
• tf.layers, tf.losses, tf.metrics -> Custom NN Models
• Core Tensorflow Python
• Core Tensorflow C++
• CPU / GPU / TPU / Android

Execution

• Code
• Tensors Definition
• Creates DAG
• Run DAG in Session
• Lazy Evaluation model (minimize context switches)

Graph and Session

• Explicit edges to represent dependencies
• Helps to partition and run parallel pieces

	import numpy as np
	import tensorflow as tf
	#numpy
	a = np.array([5,3,8])
	b = np.array([3,-1,2])
	c = np.add(a,b)
	print('numpy data')
	print(c)
	#Tensorflow
	#Build
	a = tf.constant([5,3,8])
	b = tf.constant([3,-1,2])
	c = tf.add(a,b)
	print(c)
	#scalar
	x = tf.constant(4)
	#vector
	x = tf.constant([3,5,7])
	#matrix
	#2D
	x = tf.constant([[3,5,7],[4,6,8]])
	#Get 0th column
	y = x[:,0]
	#Get the 0th row
	z = x[0,:]
	#get two values only
	#start at 0, fetch two elements
	#python range first number included last number is not
	zz = x[0,:2]
	#rewrite into 3x2 rows
	yy = tf.reshape(x,[3,2])
	#rewrite into 6x1 rows
	yyy = tf.reshape(x,[6,1])
	#3D
	x = tf.constant([[[3,5,7],[4,6,8]]])
	#placeholders to feed value
	data = tf.placeholder("float",None)
	b = data*2
	#Run
	with tf.Session() as sess:
	result = sess.run(c)
	print('tensor data')
	print(result)
	print('c evaluation')
	print(c.eval())
	print(y.eval())
	print(z.eval())
	print(zz.eval())
	print(yy.eval())
	print(yyy.eval())
	print(sess.run(b,feed_dict={data: [1,2,3]}))

view raw tensor1.py hosted with ❤ by GitHub

 
Happy Learning!!!