Skip to content

Commit ef50583

Browse files
authored
Update data preparation instructions for BioEncoder
Added instructions for arranging image dataset for training.
1 parent 0ee3fb8 commit ef50583

1 file changed

Lines changed: 17 additions & 1 deletion

File tree

help/02-data-preparation.md

Lines changed: 17 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,22 @@
11
# BioEncoder root dir
22

3-
BioEncoder uses a single working directory to store all output related to a project. The different functions and steps will create different sub-folders inside - i.e., `split_dataset` will create the _data_ folder, `train` will create _logs_, _runs_ and _weights_ folders, and `interactive_plots` will create _plots_. In the end your working directory will look like this:
3+
To train a BioEncoder model, your image dataset should be arranged like this:
4+
5+
```
6+
my-dataset/
7+
class_1/
8+
image_1.jpg
9+
image_2.jpg
10+
...
11+
class_2/
12+
image_1.jpg
13+
image_2.jpg
14+
...
15+
...
16+
```
17+
18+
19+
BioEncoder will copy them to a new working directory, where all output related to a specific project is stored. The different functions and steps will create different sub-folders inside - i.e., `split_dataset` will create the _data_ folder, `train` will create _logs_, _runs_ and _weights_ folders, and `interactive_plots` will create _plots_. In the end your working directory will look like this:
420

521
```
622
bioencoder_wd/

0 commit comments

Comments
 (0)