how to run inference/test independently
SeekPoint opened this issue · comments
I haven't implemented it yet. Since I was always testing with a given test set. I will make it available soon.
Else, you can:
dic = {entity_net.S:mb_x1,entity_net.Q:mb_x2, entity_net.keep_prob:1.0}
prediction = tf.run(entity_net.logit, feed_dict=feed_dict)
and there you have your prediction