這兩周我學習了北京大學曹建老師的TensorFlow筆記課程，認為老師講的很不錯的，很適合於想要在短期內上手完成一個相關項目的同學，課程在b站和MOOC平台都可以找到。
在卷積神經網絡一節，課程以lenet5為例，給出了完整的代碼，通過這樣一個例子完成了模型構建、較大數據量的訓練和測試。整個代碼不復雜，架構完整，我覺得代碼很干凈，很優秀，所以想把之后需要實現的Alexnet等網絡結構都按照這個代碼的結構來改。
下面是lenet5實現，數據集依然mnist。
forward.py
#coding:utf-8 import tensorflow as tf IMAGE_SIZE = 28 NUM_CHANNELS = 1 CONV1_SIZE = 5 CONV1_KERNEL_NUM = 32 CONV2_SIZE = 5 CONV2_KERNEL_NUM = 64 FC_SIZE = 512 OUTPUT_NODE = 10 def get_weight(shape, regularizer): # 參數：生成張量的維度、正則化權重 w = tf.Variable(tf.truncated_normal(shape, stddev=0.1)) if regularizer != None: tf.add_to_collection('losses', tf.contrib.layers.l2_regularizer(regularizer)(w)) return w def get_bias(shape): b = tf.Variable(tf.zeros(shape)) return b def conv2d(x, w): #參數：輸入圖片x和所用卷積核w 都為四階張量 return tf.nn.conv2d(x, w, strides=[1, 1, 1, 1], padding='SAME') def max_pool_2x2(x): return tf.nn.max_pool(x, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME') def forward(x, train, regularizer): conv1_w = get_weight([CONV1_SIZE, CONV1_SIZE, NUM_CHANNELS, CONV1_KERNEL_NUM], regularizer) conv1_b = get_bias([CONV1_KERNEL_NUM]) conv1 = conv2d(x, conv1_w) relu1 = tf.nn.relu(tf.nn.bias_add(conv1, conv1_b)) pool1 = max_pool_2x2(relu1) conv2_w = get_weight([CONV2_SIZE, CONV2_SIZE, CONV1_KERNEL_NUM, CONV2_KERNEL_NUM], regularizer) conv2_b = get_bias([CONV2_KERNEL_NUM]) conv2 = conv2d(pool1, conv2_w) relu2 = tf.nn.relu(tf.nn.bias_add(conv2, conv2_b)) pool2 = max_pool_2x2(relu2) # pool2為第二個卷積層的輸出，需要把它從三維張量變為二維張量 pool_shape = pool2.get_shape().as_list() nodes = pool_shape[1] * pool_shape[2] * pool_shape[3] # [0]是betch的值，此處我們提取[1][2][3]是特征的長、寬、深度相乘得到所有特征點的個數 reshaped = tf.reshape(pool2, [pool_shape[0], nodes]) # 重塑為二維 fcl_w = get_weight([nodes, FC_SIZE], regularizer) fcl_b = get_bias([FC_SIZE]) fcl = tf.nn.relu(tf.matmul(reshaped, fcl_w) + fcl_b) # 將二維特征輸入全連接網絡 if train: fcl = tf.nn.dropout(fcl, 0.5) # 如果是訓練階段，則對該層的輸出進行50%dropout fc2_w = get_weight([FC_SIZE, OUTPUT_NODE], regularizer) fc2_b = get_bias(OUTPUT_NODE) y = tf.matmul(fcl, fc2_w) + fc2_b return y
backward.py
#coding:utf-8 import tensorflow as tf from tensorflow.examples.tutorials.mnist import input_data import os import forward import numpy as np BATCH_SIZE = 100 LEARNING_RATE_BASE = 0.005 LEARNING_RATE_DECAY = 0.99 REGULARIZER = 0.0001 STEPS = 50000 MOVING_AVERAGE_DECAY = 0.99 MODEL_SAVE_PATH="./model/" MODEL_NAME="mnist_model" def backward(mnist): x = tf.placeholder(tf.float32, [ BATCH_SIZE, forward.IMAGE_SIZE, forward.IMAGE_SIZE, forward.NUM_CHANNELS ]) y_ = tf.placeholder(tf.float32, [None, forward.OUTPUT_NODE]) y = forward.forward(x, True, REGULARIZER) global_step = tf.Variable(0, trainable=False) ce = tf.nn.sparse_softmax_cross_entropy_with_logits(logits=y, labels=tf.arg_max(y_, 1)) cem = tf.reduce_mean(ce) loss = cem + tf.add_n(tf.get_collection('losses')) learning_rate = tf.train.exponential_decay( LEARNING_RATE_BASE, global_step, mnist.train.num_examples / BATCH_SIZE, LEARNING_RATE_DECAY, staircase=True ) train_step = tf.train.GradientDescentOptimizer(learning_rate).minimize(loss, global_step=global_step) ema = tf.train.ExponentialMovingAverage(MOVING_AVERAGE_DECAY, global_step) ema_op = ema.apply(tf.trainable_variables()) with tf.control_dependencies([train_step, ema_op]): train_op = tf.no_op(name='train') saver = tf.train.Saver() with tf.Session() as sess: init_op = tf.global_variables_initializer() sess.run(init_op) ckpt = tf.train.get_checkpoint_state(MODEL_SAVE_PATH) if ckpt and ckpt.model_checkpoint_path: saver.restore(sess, ckpt.model_checkpoint_path) for i in range(STEPS): xs, ys = mnist.train.next_batch(BATCH_SIZE) reshaped_xs = np.reshape(xs, ( BATCH_SIZE, forward.IMAGE_SIZE, forward.IMAGE_SIZE, forward.NUM_CHANNELS)) _, loss_value, step = sess.run([train_op, loss, global_step], feed_dict={x: reshaped_xs, y_: ys}) if i % 100 == 0: print("After %d training step(s), loss an training batch is %g." % (step, loss_value)) saver.save(sess, os.path.join(MODEL_SAVE_PATH, MODEL_NAME), global_step=global_step) def main(): mnist = input_data.read_data_sets("data", one_hot=True) backward(mnist) if __name__=='__main__': main()
test.py
# coding:utf-8 import time import tensorflow as tf from tensorflow.examples.tutorials.mnist import input_data import forward import backward import numpy as np TEST_INTERVAL_SECS = 5 def evaluate(mnist): with tf.Graph().as_default() as g: # 再現圖 x = tf.placeholder(tf.float32, [ mnist.test.num_examples, forward.IMAGE_SIZE, forward.IMAGE_SIZE, forward.NUM_CHANNELS]) y_ = tf.placeholder(tf.float32, [None, forward.OUTPUT_NODE]) y = forward.forward(x, False, None) # 實例化帶滑動平均的Saver對象 ema = tf.train.ExponentialMovingAverage(backward.MOVING_AVERAGE_DECAY) ema_restore = ema.variables_to_restore() saver = tf.train.Saver(ema_restore) correct_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_, 1)) accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32)) while True: with tf.Session() as sess: ckpt = tf.train.get_checkpoint_state(backward.MODEL_SAVE_PATH) # 判斷是否有模型，如果有,恢復模型到當前會話 if ckpt and ckpt.model_checkpoint_path: saver.restore(sess, ckpt.model_checkpoint_path) global_step = ckpt.model_checkpoint_path.split('/')[-1].split('-')[-1] reshaped_x = np.reshape(mnist.test.images, ( mnist.test.num_examples, forward.IMAGE_SIZE, forward.IMAGE_SIZE, forward.NUM_CHANNELS)) accuracy_score = sess.run(accuracy, feed_dict={x: reshaped_x, y_: mnist.test.labels}) print("After %s training step(s), test accuracy = %g" % (global_step, accuracy_score)) else: print('No checkpoint file found') return time.sleep(TEST_INTERVAL_SECS) def main(): mnist = input_data.read_data_sets("data", one_hot=True) evaluate(mnist) if __name__ == '__main__': main()
在自己電腦上運行還真的需要time.sleep，要不然跑起來CPU占用一直99%只能強制關機了。
while True 的循環體，會一直判斷並拿到當前最新的訓練模型，電腦上實現不能夠邊訓練邊測試，不能看到測試准確率在整個訓練過程中的變化，只能看到最后的結果啦。（訓練完成用了整整一天）
下一步就是明天參考着這個完成Alexnet的整體實現啦。