I would like to know if it is possible to do "simple" operations like adding or multiplying with a constant tensor ?
I tried this model :
tf.reset_default_graph() X = tf.placeholder(dtype=tf.float32, shape=(1,2,2,1), name = "in037") # note: I also tried to declare this one as a variable Y = tf.Variable([ [ [ , ], [ ,  ] ] ], name='testvar', dtype=tf.float32) Z = tf.math.add(Y, X, name="out037") init_op = tf.global_variables_initializer() with tf.Session() as sess: sess.run(init_op) saver = tf.train.Saver(tf.global_variables()) saver.save(sess, './test/test')
I may be doing something wrong, but I get errors, like if the compiler did not like the second variable declared. (I also checked with "--new-parser" by curiosity, but it break the assertion that a node can't have more than one producer (that feels actually weird, but I guess it's still a work in progress anyway) )
If it's not possible, where can I find a precise list of all the constraints and supported operations, etc ?
Thank you in advance !