# Train/Test/Val Split for Informer model

**URL:** <https://discuss.huggingface.co/t/train-test-val-split-for-informer-model/52994>\
**Category:** 🤗Transformers\
**Created:** [August 31, 2023, 2:30pm UTC](https://discuss.huggingface.co/t/train-test-val-split-for-informer-model/52994 "2023-08-31T14:30:46Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Christin11](https://avatars.discourse-cdn.com/v4/letter/c/919ad9/32.png) [@Christin11](https://discuss.huggingface.co/u/Christin11)\
**Post date:** [August 31, 2023, 2:30pm UTC](https://discuss.huggingface.co/t/train-test-val-split-for-informer-model/52994/1 "2023-08-31T14:30:46Z")

</div>

Thank you very much on your work on making Transformer Timeseries models easily accessible. Highly appreciated. I am currently trying to reproduce the results from the original Informer paper using the Informer Implementation from Huggingface. The authors of the original Informer paper split their datasets in 0.6/0.2/0.2 chunks. The Huggingface Informer Tutorial mentions that the val set is only “prediction length” longer than the training set and the test set only 2\* “prediction length” longer. Can I simply split the input dataset like in the original paper or do I need to adjust morein the code? Currently, I am getting results MSE/MAE on the ETHh1 dataset way better than stated in the original Informer paper.

Thank you very much in advance!

---

<div class="post-metadata">

**Author:** ![shashwat26](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/shashwat26/32/30138_2.png) [@shashwat26](https://discuss.huggingface.co/u/shashwat26)\
**Post date:** [August 3, 2024, 6:52pm UTC](https://discuss.huggingface.co/t/train-test-val-split-for-informer-model/52994/2 "2024-08-03T18:52:45Z")

</div>

Did you got your desired results ? I had created a new embedding and I want to pass it as an input to encoder; however, I am facing difficulty in this integration.  
Could you please help me?