Wanted to do a quick and dirty speed test of a tensorflow neural network model trained on the mnist data set for 25 epochs. 78.95 on the CUDA 3.0 enabled NVIDIA 680GTX tensorflow, 177 seconds for the i7 2600k @ 3.7GHZ, 396 on the I5. The 680GTX isn't a very good card for this, but it's still ~2.25x faster than an overclocked 2600k and ~5x faster than an i5 4200U on this test. The result below is for the 2600k.
In [1]:
%load_ext watermark
%watermark -a "Brett Montague" -nmv --packages numpy,statsmodels,scipy,pandas,sklearn,matplotlib,seaborn,networkx,notebook,jupyter_contrib_nbextensions
Imports and Setups¶
In [2]:
import warnings
warnings.filterwarnings('ignore')
warnings.simplefilter(action='ignore', category=FutureWarning)
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
%matplotlib inline
from sklearn import preprocessing, linear_model
from sklearn.preprocessing import scale
import sklearn.linear_model as skl_lm
from sklearn.metrics import mean_squared_error, r2_score
import statsmodels.api as sm
import statsmodels.formula.api as smf
import seaborn as sns
import tensorflow as tf
import sys
import os
from datetime import datetime, timedelta
Neural Net Code¶
In [3]:
import tensorflow as tf
import tensorflow.examples.tutorials.mnist.input_data as input_data
import os
# Shut up warning
os.environ["TF_CPP_MIN_LOG_LEVEL"]="2"
mnist = input_data.read_data_sets("/tmp/data/", one_hot=True)
n_nodes_hl1 = 500
n_nodes_hl2 = 500
n_nodes_hl3 = 500
n_classes = 10
batch_size = 100
#100 features at a time
# height x width
x = tf.placeholder('float',[None, 784])
y = tf.placeholder('float')
def neural_network_model(data):
# (input data * weights) + bias
hidden_1_layer = {'weights':tf.Variable(tf.random_normal([784, n_nodes_hl1])), 'biases':tf.Variable(tf.random_normal([n_nodes_hl1]))}
hidden_2_layer = {'weights':tf.Variable(tf.random_normal([n_nodes_hl1, n_nodes_hl2])), 'biases':tf.Variable(tf.random_normal([n_nodes_hl2]))}
hidden_3_layer = {'weights':tf.Variable(tf.random_normal([n_nodes_hl2, n_nodes_hl3])), 'biases':tf.Variable(tf.random_normal([n_nodes_hl3]))}
output_layer = {'weights':tf.Variable(tf.random_normal([n_nodes_hl3, n_classes])), 'biases':tf.Variable(tf.random_normal([n_classes]))}
l1 = tf.add(tf.matmul(data, hidden_1_layer['weights']), hidden_1_layer['biases'])
l1 = tf.nn.relu(l1)
l2 = tf.add(tf.matmul(l1, hidden_2_layer['weights']), hidden_2_layer['biases'])
l2 = tf.nn.relu(l2)
l3 = tf.add(tf.matmul(l2, hidden_3_layer['weights']), hidden_3_layer['biases'])
l3 = tf.nn.relu(l3)
output = tf.matmul(l3, output_layer['weights']) + output_layer['biases']
return output
def train_neural_network(x):
prediction = neural_network_model(x)
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=prediction,labels=y))
optimizer = tf.train.AdamOptimizer().minimize(cost)
#epochs = cycles of feed forward + backprop
hm_epochs = 25
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
for epoch in range(hm_epochs):
epoch_loss = 0
for _ in range(int(mnist.train.num_examples/batch_size)):
epoch_x, epoch_y = mnist.train.next_batch(batch_size)
_, c = sess.run([optimizer, cost], feed_dict={x: epoch_x, y: epoch_y})
epoch_loss += c
print('Epoch:', epoch, ' completed out of ', hm_epochs, ' loss: ', epoch_loss)
correct = tf.equal(tf.argmax(prediction,1), tf.argmax(y,1))
accuracy = tf.reduce_mean(tf.cast(correct,'float'))
print('Accuracy:', accuracy.eval({x:mnist.test.images, y:mnist.test.labels}))
import time
start_time = time.time()
train_neural_network(x)
print("--- %s seconds ---" % (time.time() - start_time))
In [ ]: