Keras Tensorboard callback not writing images -
i trying visualize weights of keras model tensorboard. here model using:
model = sequential([ conv2d(filters=32, kernel_size=(3,3), padding="same", activation='relu', input_shape=(40,40,3)), maxpooling2d(pool_size=(2, 2)), conv2d(filters=64, kernel_size=(5,5), padding="same", activation='relu'), maxpooling2d(pool_size=(2, 2)), flatten(), dense(1024, activation='relu'), dropout(0.5), dense(43, activation='softmax'), ]) model.compile(optimizer='sgd', loss='categorical_crossentropy', metrics=['accuracy'])
and training call:
model.fit_generator( ... callbacks = [ modelcheckpoint('models/gtsrb1-{epoch}.hdf5', verbose=1, save_best_only = true), tensorboard(log_dir='tblogs/', write_graph=true, write_grads=true, write_images=true), earlystopping(patience=5, verbose=1), ],)
however, when start tensorboard, get:
scalars , graphs looks okay not problem of wrong logdir
. doing wrong here?
you need add histogram_freq=x
, x
should different zero, writing of images enabled.
but if this, might still fail, depending on version of keras (see https://github.com/fchollet/keras/issues/6096)
Comments
Post a Comment