From a0742fd0eaceae0ae611eaa77c4fb64f94974f3c Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz@telenczuk.pl>
Date: Tue, 15 Oct 2019 21:06:22 +0200
Subject: [PATCH] doc: explain mechanism of validation data split

---
 keras_preprocessing/image/image_data_generator.py | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/keras_preprocessing/image/image_data_generator.py b/keras_preprocessing/image/image_data_generator.py
index 9a2bb38b..1fd26102 100644
--- a/keras_preprocessing/image/image_data_generator.py
+++ b/keras_preprocessing/image/image_data_generator.py
@@ -504,7 +504,8 @@ def flow_from_directory(self,
             follow_links: Whether to follow symlinks inside
                 class subdirectories (default: False).
             subset: Subset of data (`"training"` or `"validation"`) if
-                `validation_split` is set in `ImageDataGenerator`.
+                `validation_split` is set in `ImageDataGenerator`. Files
+                will be split after sorting in alphanumeric order.
             interpolation: Interpolation method used to
                 resample the image if the
                 target size is different from that of the loaded image.