Eval method seems to be using data from the train dataset

santhoshkolloju / Abstractive-Summarization-With-Transfer-Learning

Abstractive summarisation using Bert as encoder and Transformer Decoder

407 stars 98 forks source link

Eval method seems to be using data from the train dataset #7

Closed Vibha111094 closed 5 years ago

Vibha111094 commented 5 years ago

I observed that although I use feed_dict = { iterator.handle: iterator.get_handle(sess, 'eval'), tx.global_mode(): tf.estimator.ModeKeys.EVAL, } in the _eval_epoch method ,I observed that it is using a few examples from train dataset as well. Is this the desired behavior as we use FeedableDataIterator which is supposed to iterates through multiple datasets and switches between datasets.

If so could you please explain why such a behavior is necessary .

santhoshkolloju commented 5 years ago

iterator.get_handle(sess, 'eval') will get the handle for Eval data set. It will not run on train data set