2
votes

Hi I'm using Keras to train a neural network. I need to use tensorboard to track how my training goes. From following documents, I used following code to use tensorflow in Keras.

tensorboard = TensorBoard(log_dir='./logs', histogram_freq=0,
                      write_graph=True, write_images=False)

hist = custom_resnet_model.fit(X_train, X_valid, batch_size=32, epochs=nb_epoch, verbose=1, validation_data=(Y_train, Y_valid),callbacks=[tensorboard])

And my 'graph' folder has got following file as well.

events.out.tfevents.1520742430.HIR

But when trying to visualize this, using command

tensorboard --logdir=/C:/CT_SCAN_IMAGE_SET/resnet_50/dbs2017/logs

enter image description here

Can someone please help me to identify the steps that I have missed

1
I assume you are using windows, have you tried without the first / in the log path ? Using /C: doesn't work for me while using `C:\ does. - kluu
Ohhh it is working... :) Thanks alot dear friend... Please put this as an answer. So I can accept it. Million thanks for you... - user3789200
Haha happens to me all the time. Ok cool :D - kluu

1 Answers

2
votes

Change tensorboard --logdir=/C:/CT_SCAN_IMAGE_SET/resnet_50/dbs2017/logs to tensorboard --logdir=C:/CT_SCAN_IMAGE_SET/resnet_50/dbs2017/logs, so you can train your model peacefully.