Python – How do I display a list of all gradients in Tensorboard?

I’m using Tensorboard 1.5 and I want to see how my gradients look.

Here is an example of the layer I’m using:

net = tf.layers.dense(features, 40, activation=tf.nn.relu, kernel_regularizer=regularizer, 
                        kernel_initializer=tf.contrib.layers.xavier_initializer())

Here is my optimizer:

train_op = tf.train.AdamOptimizer(learning_rate = learning_rate).minimize(loss)

For my model parameters, I create the summary this way:

for var in tf.trainable_variables():
    tf.summary.histogram(var.name, var)

Is there a similar way to get all gradients in the for loop to create my summary?

You should first use the optimizer’s compute_gradients to get gradients and then pass them to summary:

opt = tf.train.AdamOptimizer(learning_rate = learning_rate)

# Calculate the gradients for the batch of data
grads = opt.compute_gradients(loss)

# Add histograms for gradients.
for grad, var in grads:
  if grad is not None:
    summaries.append(tf.summary.histogram(var.op.name + '/gradients', grad))

Then for training, you can call the optimizer’s apply_gradients:

# Apply the gradients to adjust the shared variables.
train_op = opt.apply_gradients(grads, global_step=global_step)

More can be found in tensorflow cifar10 tutorial .