memory usage

[s3xton]

Is this model particularly memory intensive? The model hangs on building the loss model for seq_length 10 having used 16GB of RAM.

This might be caused by the fact that I am using tensorflow 1.0.0-rc1 (because of #26, legacy_seq2seq is present in 1.0.0-rc1 and not 0.12.1). I've modified your code to run on this version, and it trains correctly with lower length sequences, but still eats memory.

Is this normal, or some sort of memory leak caused by the newer version of tensorflow?

[henghuiz-zz]

I think the issue is this code saves one individual model for every possible sequence length. So it will take a huge amount of memory. I was trying to improve it but I then soon I find I don't need NTM and gave up.

[s3xton]

Ah I see thank you. Do you have any suggestions as to how to approach solving this issue?

[camigord]

Hi @s3xton, check this implementation I have recently developed following Mostafa-Samir's approach.