# Composition Training/Validation Split of AutoTrain

**URL:** </static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fcomposition-training-validation-split-of-autotrain%2F26328%3E%5C
**Category:** Research\
**Created:** [November 18, 2022, 7:04pm UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fcomposition-training-validation-split-of-autotrain%2F26328 "2022-11-18T19:04:20Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![ropodo](https://avatars.discourse-cdn.com/v4/letter/r/ecb155/32.png) [@ropodo](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2Fropodo)%5C
**Post date:** [November 18, 2022, 7:04pm UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fcomposition-training-validation-split-of-autotrain%2F26328%2F1 "2022-11-18T19:04:20Z")

</div>

Hey everyone,

is there any documentation about how AutoTrain splits your data into training and validation data? I selected the option that it should automatically do so. I conducted binary text classification with BERT.  
It would be great to know in order to report a percentage split in a research project.

Thanks!

Best

rob
