tensorflow - Tensorboard scalars and graphs duplicated -


i'm using tensorboard visualize network metrics , graph.

i create session sess = tf.interactivesession() , build graph in jupyter notebook.

in graph, include 2 summary scalars:

with tf.variable_scope('summary') scope:        loss_summary = tf.summary.scalar('loss', cross_entropy)    train_accuracy_summary = tf.summary.scalar('train_accuracy', accuracy)

i create summary_writer = tf.summary.filewriter(logdir, sess.graph) , run:

_,loss_sum,train_accuracy_sum=sess.run([...],feed_dict=feed_dict)

i write metrics:

summary_writer.add_summary(loss_sum, i) summary_writer.add_summary(train_accuracy_sum, i)

i run code 3 times.

each time run, re-import tf , create new interactive session.

but, in tensorboard, separate scalar window created each run:

enter image description here

also, graph appears duplicated if check data last run:

enter image description here

how prevent duplication of graph , scalar window each time run?

  • i want data appear in same scalar plots (with multiple series / plot).
  • i want each run reference single graph visualization.

i suspect problem arises because running code 3 times in process (same script, jupyter notebook, or whatever), , invocations share same "default graph" in tensorflow. tensorflow needs give each node in graph unique name, appends "_1" , "_2" names of summary nodes in second , third invocations.

how avoid this? easiest way create new graph each time run code. there (at least) 3 ways this:

  • wrap code in with tf.graph().as_default(): block, constructs new tf.graph object , sets default graph extent of with block.

  • if construct session before creating graph, can construct session sess = tf.interactivesession(graph=tf.graph()). newly constructed tf.graph object remains default graph until call sess.close().

  • call tf.reset_default_graph() between invocations of code.

the with-block approach "most structured" way things, , might best if writing standalone script. however, since using tf.interactivesession, assume using interactive repl of kind, , other 2 approaches more useful (e.g. splitting execution across multiple cells).


Comments

Post a Comment

Popular posts from this blog

networking - Vagrant-provisioned VirtualBox VM is not reachable from Ubuntu host -

this question exact duplicate of: puphpet/vagrant: unable access lamp vm ip address 1 answer i following basic vagrant tutorial , vagrant provisions box on ubuntu without complaining. moreover, can vagrant ssh vm. however, vm not reachable host: it's not possible ping it's ip address. what have achieved: installed virtualbox (5.1.24 r117012 (qt5.5.1)) , vagrant (1.9.7) onto ubuntu (16.04.2 lts) machine. provisioned vm vagrant up , vagrant finishes job , not complain. vagrant ssh works. i have manually installed lighttpd on guest vm , can ping/curl address: localhost within guest vm. this when trying reach guest vm host: ~/tmp/myfirstvagrantproject$ ping 172.22.22.22 ping 172.22.22.22 (172.22.22.22) 56(84) bytes of data. ... icmp_seq=4 destination net unreachable ... 171 packets transmitted, 0 received, +21 errors, 100% packet loss, time 173
Read more

c# - ASP.NET Core - There is already an object named 'AspNetRoles' in the database -

i'm receiving error applicationdbcontext database when trying update entity framework. i've googled, tried fixes here on so, no avail. i'm deleting database each time make change, once deploy, not option. here's error: pm> update-database -context applicationdbcontext system.data.sqlclient.sqlexception: there object named 'aspnetroles' in database. @ system.data.sqlclient.sqlconnection.onerror(sqlexception exception, boolean breakconnection, action 1 wrapcloseinaction) @ system.data.sqlclient.tdsparser.throwexceptionandwarning(tdsparserstateobject stateobj, boolean callerhasconnectionlock, boolean asyncclose) @ system.data.sqlclient.tdsparser.tryrun(runbehavior runbehavior, sqlcommand cmdhandler, sqldatareader datastream, bulkcopysimpleresultset bulkcopyhandler, tdsparserstateobject stateobj, boolean& dataready) @ system.data.sqlclient.sqlcommand.runexecutenonquerytds(string methodname, boolean async, int32 timeout, boolean
Read more

ruby on rails - ArgumentError: Missing host to link to! Please provide the :host parameter, set default_url_options[:host], or set :only_path to true -

i have built module in solidus , ruby on rails , try create tests. i trying create integration tests , have error: e error: prescriptionphotostest#test_/prescription_photos_post_endpoint_throws_an_error_when_wrong_params: argumenterror: missing host link to! please provide :host parameter, set default_url_options[:host], or set :only_path true test/integration/prescription_photos_test.rb:6:in `block in ' i read this , this , doesn't make difference. my code returning error on line: @user = factorygirl.create(:user) here code files. prescription_photos_rest.rb require 'test_helper' class prescriptionphotostest < actiondispatch::integrationtest setup spree::auth::config[:confirmable] = false @user = factorygirl.create(:user) end def authenticated_header token = knock::authtoken.new(payload: { sub: @user.id }).token { 'authorization': "bearer #{token}" } end test '/prescription_p
Read more