Exhaust training set examples per epoch

Author: eywalker

02_logistic_regression.py

@@ -33,7 +33,7 @@ def model(X, w):
     tf.initialize_all_variables().run()
 
     for i in range(100):
-        for start, end in zip(range(0, len(trX), 128), range(128, len(trX), 128)):
+        for start, end in zip(range(0, len(trX), 128), range(128, len(trX)+1, 128)):
             sess.run(train_op, feed_dict={X: trX[start:end], Y: trY[start:end]})
         print(i, np.mean(np.argmax(teY, axis=1) ==
                          sess.run(predict_op, feed_dict={X: teX, Y: teY})))

03_net.py

@@ -35,7 +35,7 @@ def model(X, w_h, w_o):
     tf.initialize_all_variables().run()
 
     for i in range(100):
-        for start, end in zip(range(0, len(trX), 128), range(128, len(trX), 128)):
+        for start, end in zip(range(0, len(trX), 128), range(128, len(trX)+1, 128)):
             sess.run(train_op, feed_dict={X: trX[start:end], Y: trY[start:end]})
         print(i, np.mean(np.argmax(teY, axis=1) ==
                          sess.run(predict_op, feed_dict={X: teX, Y: teY})))

04_modern_net.py

@@ -45,7 +45,7 @@ def model(X, w_h, w_h2, w_o, p_keep_input, p_keep_hidden): # this network is the
     tf.initialize_all_variables().run()
 
     for i in range(100):
-        for start, end in zip(range(0, len(trX), 128), range(128, len(trX), 128)):
+        for start, end in zip(range(0, len(trX), 128), range(128, len(trX)+1, 128)):
             sess.run(train_op, feed_dict={X: trX[start:end], Y: trY[start:end],
                                           p_keep_input: 0.8, p_keep_hidden: 0.5})
         print(i, np.mean(np.argmax(teY, axis=1) ==

05_convolutional_net.py

@@ -66,7 +66,7 @@ def model(X, w, w2, w3, w4, w_o, p_keep_conv, p_keep_hidden):
 
     for i in range(100):
         training_batch = zip(range(0, len(trX), batch_size),
-                             range(batch_size, len(trX), batch_size))
+                             range(batch_size, len(trX)+1, batch_size))
         for start, end in training_batch:
             sess.run(train_op, feed_dict={X: trX[start:end], Y: trY[start:end],
                                           p_keep_conv: 0.8, p_keep_hidden: 0.5})

06_autoencoder.py

@@ -50,7 +50,7 @@ def model(X, mask, W, b, W_prime, b_prime):
     tf.initialize_all_variables().run()
 
     for i in range(100):
-        for start, end in zip(range(0, len(trX), 128), range(128, len(trX), 128)):
+        for start, end in zip(range(0, len(trX), 128), range(128, len(trX)+1, 128)):
             input_ = trX[start:end]
             mask_np = np.random.binomial(1, 1 - corruption_level, input_.shape)
             sess.run(train_op, feed_dict={X: input_, mask: mask_np})

09_tensorboard.py

@@ -60,7 +60,7 @@ def model(X, w_h, w_h2, w_o, p_keep_input, p_keep_hidden):
     tf.initialize_all_variables().run()
 
     for i in range(100):
-        for start, end in zip(range(0, len(trX), 128), range(128, len(trX), 128)):
+        for start, end in zip(range(0, len(trX), 128), range(128, len(trX)+1, 128)):
             sess.run(train_op, feed_dict={X: trX[start:end], Y: trY[start:end],
                                           p_keep_input: 0.8, p_keep_hidden: 0.5})
         summary, acc = sess.run([merged, acc_op], feed_dict={X: teX, Y: teY,

10_save_restore_net.py

@@ -70,7 +70,7 @@ def model(X, w_h, w_h2, w_o, p_keep_input, p_keep_hidden): # this network is the
     print("Start from:", start)
 
     for i in range(start, 100):
-        for start, end in zip(range(0, len(trX), 128), range(128, len(trX), 128)):
+        for start, end in zip(range(0, len(trX), 128), range(128, len(trX)+1, 128)):
             sess.run(train_op, feed_dict={X: trX[start:end], Y: trY[start:end],
                                           p_keep_input: 0.8, p_keep_hidden: 0.5})