Question

我在Jupyter笔记本中执行以下代码

import tensorflow as tf
# Parameters
learning_rate = 0.01
training_epochs = 25
display_step = 1

# tf Graph Input
x = tf.placeholder("float", [None, dims]) 
y = tf.placeholder("float", [None, nb_classes])

# Construct (linear) model
with tf.name_scope("model") as scope:
    # Set model weights
    W = tf.Variable(tf.zeros([dims, nb_classes]))
    b = tf.Variable(tf.zeros([nb_classes]))
    activation = tf.nn.softmax(tf.matmul(x, W) + b) # Softmax

    # Add summary ops to collect data
    w_h = tf.summary.histogram("weights_histogram", W)
    b_h = tf.summary.histogram("biases_histograms", b)
    tf.summary.scalar('mean_weights', tf.reduce_mean(W))
    tf.summary.scalar('mean_bias', tf.reduce_mean(b))

# Minimize error using cross entropy
# Note: More name scopes will clean up graph representation
with tf.name_scope("cost_function") as scope:
    cross_entropy = y*tf.log(activation)
    cost = tf.reduce_mean(-tf.reduce_sum(cross_entropy,reduction_indices=1))
    # Create a summary to monitor the cost function
    tf.summary.scalar("cost_function", cost)
    tf.summary.histogram("cost_histogram", cost)

with tf.name_scope("train") as scope:
    # Set the Optimizer
    optimizer = tf.train.GradientDescentOptimizer(learning_rate).minimize(cost)

with tf.name_scope('Accuracy') as scope:
    correct_prediction = tf.equal(tf.argmax(activation, 1), tf.argmax(y, 1))
    # Calculate accuracy
    accuracy = tf.reduce_mean(tf.cast(correct_prediction, "float"))
    # Create a summary to monitor the cost function
    tf.summary.scalar("accuracy", accuracy)

LOGDIR = '/tmp/logistic_logs'
import os, shutil
if os.path.isdir(LOGDIR):
    shutil.rmtree(LOGDIR)
os.mkdir(LOGDIR)

# Plug TensorBoard Visualisation 
writer = tf.summary.FileWriter(LOGDIR, graph=tf.get_default_graph())

for var in tf.get_collection(tf.GraphKeys.SUMMARIES):
    print(var.name)

summary_op = tf.summary.merge_all()
print('Summary Op: ' + summary_op)


# Launch the graph
with tf.Session() as session:
    # Initializing the variables
    session.run(tf.global_variables_initializer())

    cost_epochs = []
    # Training cycle
    for epoch in range(training_epochs):
        _, summary, c = session.run(fetches=[optimizer, summary_op, cost], 
                                    feed_dict={x: X_train, y: Y_train})
        cost_epochs.append(c)
        writer.add_summary(summary=summary, global_step=epoch)
        print("accuracy epoch {}:{}".format(epoch, accuracy.eval({x: X_train, y: Y_train})))

    print("Training phase finished")

    #plotting
    plt.plot(range(len(cost_epochs)), cost_epochs, 'o', label='Logistic Regression Training phase')
    plt.ylabel('cost')
    plt.xlabel('epoch')
    plt.legend()
    plt.show()

    prediction = tf.argmax(activation, 1)
    print(prediction.eval({x: X_test}))


%%bash
tensorboard --logdir=/tmp/logistic_logs

当我执行tensorboard命令时，我没有得到任何标量，图形等。它只是加载tensorboard仪表板。例如，对于图，我得到："No graph definition files were found."

Answer 1

我在Jupyter Notebook所在的目录中手动创建了一个名为logistic_logs的目录，然后运行以下代码。这对我有用：

import tensorflow as tf
# Parameters
learning_rate = 0.01
training_epochs = 25
display_step = 1
dims = 3
nb_classes = 2

tf.reset_default_graph()

# tf Graph Input
x = tf.placeholder("float", [None, dims]) 
y = tf.placeholder("float", [None, nb_classes])

# Construct (linear) model
with tf.name_scope("model") as scope:
    # Set model weights
    W = tf.Variable(tf.zeros([dims, nb_classes]))
    b = tf.Variable(tf.zeros([nb_classes]))
    activation = tf.nn.softmax(tf.matmul(x, W) + b) # Softmax

    # Add summary ops to collect data
    w_h = tf.summary.histogram("weights_histogram", W)
    b_h = tf.summary.histogram("biases_histograms", b)
    tf.summary.scalar('mean_weights', tf.reduce_mean(W))
    tf.summary.scalar('mean_bias', tf.reduce_mean(b))

# Minimize error using cross entropy
# Note: More name scopes will clean up graph representation
with tf.name_scope("cost_function") as scope:
    cross_entropy = y*tf.log(activation)
    cost = tf.reduce_mean(-tf.reduce_sum(cross_entropy,reduction_indices=1))
    # Create a summary to monitor the cost function
    tf.summary.scalar("cost_function", cost)
    tf.summary.histogram("cost_histogram", cost)

with tf.name_scope("train") as scope:
    # Set the Optimizer
    optimizer = tf.train.GradientDescentOptimizer(learning_rate).minimize(cost)

with tf.name_scope('Accuracy') as scope:
    correct_prediction = tf.equal(tf.argmax(activation, 1), tf.argmax(y, 1))
    # Calculate accuracy
    accuracy = tf.reduce_mean(tf.cast(correct_prediction, "float"))
    # Create a summary to monitor the cost function
    tf.summary.scalar("accuracy", accuracy)

# Plug TensorBoard Visualisation 
writer = tf.summary.FileWriter("logistic_logs", graph=tf.get_default_graph())

for var in tf.get_collection(tf.GraphKeys.SUMMARIES):
    print(var.name)

summary_op = tf.summary.merge_all()
print('Summary Op: ' + summary_op)


# Launch the graph
with tf.Session() as session:
    # Initializing the variables
    session.run(tf.global_variables_initializer())
    writer.add_graph(session.graph)

请注意，问题可能在于您在构建图形之前未重置默认图形。第一次执行代码时，便已经构建了默认图形，因此不要忘记在构建默认图形之前总是重置它。我的代码中的这一行处理以下问题：

tf.reset_default_graph()

然后，运行：

tensorboard --logdir=logistic_logs/

TensorBoard不启动任何图形

1 个答案: