Calling TensorBoard from Keras to inspect diagnostics - python-3.x

Question

I am trying to use TensorBoard with Keras and I follow the instructions in this short tutorial: http://fizzylogic.nl/2017/05/08/monitor-progress-of-your-keras-based-neural-network-using-tensorboard/
My code is:
from time import time
from keras.models import Sequential
from keras.layers import Dense, Activation
from keras.callbacks import TensorBoard
# Compile the model
model.compile(optimizer = 'adam', loss = 'categorical_crossentropy', metrics = ['accuracy'])
tensorboard = TensorBoard(log_dir="logs/{}".format(time()))
# Fit the model
model.fit(X, y, validation_split = 0.3, epochs=30, callbacks = [tensorboard])
The code is executed without any problem.
Following the advice of the tutorial:
Monitoring progress Now that you have a tensorboard instance hooked up
you can start to monitor the program by executing the following
command in a separate terminal:
tensorboard --logdir=logs/
I open a terminal and execute the aforementioned command. This is what I get:
(base) C:\Users\Alienware\Documents>tensorboard --logdir=logs/
2019-01-07 22:02:56.109894: I tensorflow/core/platform/cpu_feature_guard.cc:140] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX AVX2
TensorBoard 1.8.0 at http://ALIENWARE-PC:6006 (Press CTRL+C to quit)
W0107 22:04:55.763794 Thread-1 application.py:274] path /[[_dataImageSrc]] not found, sending 404
W0107 22:04:55.779416 Thread-1 application.py:274] path /[[_imageURL]] not found, sending 404
I then open the webpage. This is what I see:
How can I sort this out?

Answer 1

It needs to the exact folder that the data is saved into. You format log_dir="logs/{}".format(time()) the directory using the current which is obviously not the same directory as logs/. By default the logs directory is keras.callbacks.TensorBoard(log_dir='./logs',...).
You would need to either remove the time formatting or start tensorboard on the correct directory.