daily pastebin goal
76%
SHARE
TWEET

Untitled

a guest Mar 18th, 2019 63 Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
  1. import tensorflow as tf
  2. from tensorflow.examples.tutorials.mnist import input_data
  3. mnist = input_data.read_data_sets("/tmp/data/", one_hot = True)
  4.  
  5. batch_size = 100
  6.  
  7. #MODEL
  8.  
  9. def neural_network_model(data):
  10.     weights = {'W_conv': tf.Variable(tf.random_normal([5,5,1,32])),
  11.                'W_fc': tf.Variable(tf.random_normal([14*14*32,100])),
  12.                'W_out': tf.Variable(tf.random_normal([100,10])),}
  13.    
  14.    
  15.     biases = {'b_conv':tf.Variable(tf.random_normal([32])),
  16.               'b_fc':tf.Variable(tf.random_normal([100])),
  17.               'b_out':tf.Variable(tf.random_normal([10])),
  18.              }
  19.  
  20.     x = tf.reshape(data,shape=[-1,28,28,1])
  21.  
  22.     c = tf.nn.conv2d(x,weights['W_conv'], strides = [1,1,1,1],padding= 'SAME')
  23.     c += biases["b_conv"]
  24.     c = tf.nn.relu(c)
  25.    
  26.     m = tf.nn.max_pool(c,ksize=[1,2,2,1], strides = [1,2,2,1], padding = "SAME")
  27.    
  28.    
  29.     fc = tf.reshape(m,[-1,14*14*32])
  30.     fc = tf.nn.relu(tf.add(tf.matmul(fc,weights['W_fc']),biases["b_fc"]))
  31.    
  32.     output = tf.nn.relu(tf.matmul(fc,weights['W_out'])+biases["b_out"])
  33.    
  34.    
  35.     return output
  36.  
  37.  
  38. #GRAF  
  39.  
  40. x = tf.placeholder('float',[None,784])
  41. y = tf.placeholder('float')
  42.  
  43. predictions = neural_network_model(x)
  44. cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits_v2(logits=predictions,labels=y))
  45.  
  46. optimizer = tf.train.AdamOptimizer().minimize(cost)
  47.  
  48. saver = tf.train.Saver()
  49.  
  50. # WYWOŁANIE SESJI
  51.  
  52. hm_epochs =3
  53.  
  54. with tf.Session() as sess:
  55.  
  56.     sess.run(tf.global_variables_initializer())
  57.  
  58.     for epoch in range(hm_epochs):
  59.  
  60.         epoch_loss =0
  61.         for _ in range(int(mnist.train.num_examples/batch_size)):
  62.             epoch_x,epoch_y = mnist.train.next_batch(batch_size)
  63.             _, c = sess.run([optimizer,cost],feed_dict = {x:epoch_x,y:epoch_y})
  64.             epoch_loss +=c
  65.         print("Epoch",epoch,'completed out of',hm_epochs,'loss:', epoch_loss)
  66.  
  67.        
  68.  
  69.     correct = tf.equal(tf.argmax(predictions,1), tf.argmax(y,1))
  70.     accuracy = tf.reduce_mean(tf.cast(correct,'float'))
  71.     print("Accuracy:",accuracy.eval({x:mnist.train.images, y:mnist.train.labels}))
  72.  
  73.    
  74.     save_path = saver.save(sess, "/tmp/model.ckpt")
RAW Paste Data
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy. OK, I Understand
 
Top