Keras: get warnings for tensorboard : summary name is illegal

[hanzlfs]

After I add tensorboard into callbacks, it shows warning of summary name format is illegal. I suspect it could be caused by the keras model summary format.

warnings

INFO:tensorflow:Summary name convolution2d_1_W:0 is illegal; using convolution2d_1_W_0 instead.

code

tensorboard = keras.callbacks.TensorBoard(
    log_dir='tensorboard/esc_example/' + name + '/', 
    write_graph=True,
    histogram_freq=1
)

[Geeshang]

Got same problems, when add TensorBoard callback to the example https://github.com/fchollet/keras/blob/master/examples/cifar10_cnn.py

tensorflow (1.0.1)
Keras (2.0.0)

INFO:tensorflow:Summary name conv2d_1/kernel:0 is illegal; using conv2d_1/kernel_0 instead.
INFO:tensorflow:Summary name conv2d_1/kernel:0 is illegal; using conv2d_1/kernel_0 instead.
INFO:tensorflow:Summary name conv2d_1/bias:0 is illegal; using conv2d_1/bias_0 instead.
INFO:tensorflow:Summary name conv2d_1/bias:0 is illegal; using conv2d_1/bias_0 instead.
INFO:tensorflow:Summary name conv2d_2/kernel:0 is illegal; using conv2d_2/kernel_0 instead.
INFO:tensorflow:Summary name conv2d_2/kernel:0 is illegal; using conv2d_2/kernel_0 instead.
INFO:tensorflow:Summary name conv2d_2/bias:0 is illegal; using conv2d_2/bias_0 instead.
INFO:tensorflow:Summary name conv2d_2/bias:0 is illegal; using conv2d_2/bias_0 instead.
INFO:tensorflow:Summary name conv2d_3/kernel:0 is illegal; using conv2d_3/kernel_0 instead.
INFO:tensorflow:Summary name conv2d_3/kernel:0 is illegal; using conv2d_3/kernel_0 instead.
INFO:tensorflow:Summary name conv2d_3/bias:0 is illegal; using conv2d_3/bias_0 instead.
INFO:tensorflow:Summary name conv2d_3/bias:0 is illegal; using conv2d_3/bias_0 instead.
INFO:tensorflow:Summary name conv2d_4/kernel:0 is illegal; using conv2d_4/kernel_0 instead.
INFO:tensorflow:Summary name conv2d_4/kernel:0 is illegal; using conv2d_4/kernel_0 instead.
INFO:tensorflow:Summary name conv2d_4/bias:0 is illegal; using conv2d_4/bias_0 instead.
INFO:tensorflow:Summary name conv2d_4/bias:0 is illegal; using conv2d_4/bias_0 instead.
INFO:tensorflow:Summary name dense_1/kernel:0 is illegal; using dense_1/kernel_0 instead.
INFO:tensorflow:Summary name dense_1/kernel:0 is illegal; using dense_1/kernel_0 instead.
INFO:tensorflow:Summary name dense_1/bias:0 is illegal; using dense_1/bias_0 instead.
INFO:tensorflow:Summary name dense_1/bias:0 is illegal; using dense_1/bias_0 instead.
INFO:tensorflow:Summary name dense_2/kernel:0 is illegal; using dense_2/kernel_0 instead.
INFO:tensorflow:Summary name dense_2/kernel:0 is illegal; using dense_2/kernel_0 instead.
INFO:tensorflow:Summary name dense_2/bias:0 is illegal; using dense_2/bias_0 instead.
INFO:tensorflow:Summary name dense_2/bias:0 is illegal; using dense_2/bias_0 instead.

[cattaneod]

I have the same issue.

[liyi193328]

same issue, this error results to the variable's summary can't display.

[lk1983823]

I have the same issue

[codeisnotcode]

Same issue when doing this to track shifts in trainable parameters during training by monitoring their mean value:

for v in tf.trainable_variables():
   vrm = tf.reduce_mean(v)
   tf.summary.scalar(v.name, vrm)

[HappyStorm]

Same issue here

[hipoglucido]

I am having also this issue. But I only get that when I am running on Jupyter notebook. I don't see any of that while running from a .py script

[sivagnanamn]

Same issue here

[studynon]

same issues in tensorflow, not only in keras.

[SeaFX]

The same issue, sorted out by replacing variable.name with variable.name.replace(':','_')

The summary names do not like :

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed after 30 days if no further activity occurs, but feel free to re-open a closed issue if needed.

[erichiller]

Yes, this is a tensorflow issue nearly certain.

[post2web]

Don't add spaces or other special chars in the summary name. For example instead of:

tf.summary.scalar('classify loss', classify_loss)

do:

tf.summary.scalar('classify_loss', classify_loss)

[wt-huang]

Closing as this is resolved