# Is 1 week of course mostely inactual with new tf versions?

**URL:** <https://community.deeplearning.ai/t/is-1-week-of-course-mostely-inactual-with-new-tf-versions/430760>\
**Category:** Natural Language Processing in TensorFlow\
**Tags:** week-module-1\
**Created:** [August 28, 2023, 8:16pm UTC](https://community.deeplearning.ai/t/is-1-week-of-course-mostely-inactual-with-new-tf-versions/430760 "2023-08-28T20:16:01Z")\
**Posts on this page:** 1\
**Showing post:** 4

<div class="post-metadata">

**Author:** ![balaji.ambresh](https://sea1.discourse-cdn.com/dlai/user_avatar/community.deeplearning.ai/balaji.ambresh/32/128697_2.png) [@balaji.ambresh](https://community.deeplearning.ai/u/balaji.ambresh)\
**Post date:** [August 30, 2023, 8:01am UTC](https://community.deeplearning.ai/t/is-1-week-of-course-mostely-inactual-with-new-tf-versions/430760/4 "2023-08-30T08:01:11Z")

</div>

Performance is one of the main reasons for deprecation of `tf.keras.preprocessing` layers.

For instance, when using `image_dataset_from_directory`, augmentations are done in layers within the model which internally uses GPU. `ImageDataGenerator` uses CPU to perform augmentations prior to feeding data into the model.

I’ve observed a good amount of difference when it comes to training time when using the newer recommended APIs.

---

_[View the full topic](https://community.deeplearning.ai/t/is-1-week-of-course-mostely-inactual-with-new-tf-versions/430760)._