# NPE when load TF to SameDiff **URL:** \ **Category:** SameDiff\ **Created:** [October 26, 2021, 8:37am UTC](https://community.konduit.ai/t/npe-when-load-tf-to-samediff/1661 "2021-10-26T08:37:44Z")\ **Posts on this page:** 3\ **Page:** 2
**Author:** ![agibsonccc](https://yyz1.discourse-cdn.com/flex035/user_avatar/community.konduit.ai/agibsonccc/32/697_2.png) [@agibsonccc](https://community.konduit.ai/u/agibsonccc)\ **Post date:** [December 3, 2021, 11:58am UTC](https://community.konduit.ai/t/npe-when-load-tf-to-samediff/1661/21 "2021-12-03T11:58:45Z")
@SidneyLann no please reread my post. I’m saying _samediff_ now has an eager mode that runs during model import. This uses more memory because we store intermediate results while the graph is executing. ---
**Author:** ![SidneyLann](https://yyz1.discourse-cdn.com/flex035/user_avatar/community.konduit.ai/sidneylann/32/367_2.png) [@SidneyLann](https://community.konduit.ai/u/SidneyLann)\ **Post date:** [December 3, 2021, 12:04pm UTC](https://community.konduit.ai/t/npe-when-load-tf-to-samediff/1661/22 "2021-12-03T12:04:33Z")
Does SameDiff has an eager mode like that in tf2.x? Or they are different things? ---
**Author:** ![agibsonccc](https://yyz1.discourse-cdn.com/flex035/user_avatar/community.konduit.ai/agibsonccc/32/697_2.png) [@agibsonccc](https://community.konduit.ai/u/agibsonccc)\ **Post date:** [December 3, 2021, 12:10pm UTC](https://community.konduit.ai/t/npe-when-load-tf-to-samediff/1661/23 "2021-12-03T12:10:51Z")
@SidneyLann it’s mainly used for model import right now. Think of it as a half eager mode that we want to extend to allow for general use. There are some limitations right now like not being properly integrated in to the multi threaded sessions etc where normal arrays are stored. Eager mode computed op outputs have their own separate storage that is specifically built for model import use cases. The beginnings of a real eager mode are there though yes. [Previous page](https://community.konduit.ai/t/npe-when-load-tf-to-samediff/1661.md?page=1)