Generate configure for YOLO
Description
- Automatic Generate configure files to train YOLO.
- All you have to do is prepare the files for the dataset and labels, and the files needed for training will be generated.
- You may need to prepare the yolov3.cfg file and customize it yourself. The weight uses darknet53.conv.74.
Requirement
- Ubuntu18.04LTS
- darknet YOLO v3(v4) https://github.com/pjreddie/darknet/tree/yolov3
Installation
git clone https://github.com/Ar-Ray-code/generate_cfg_for_yolo
Usage
sh generate_config.sh <target_folder_path> <target_data_folder_path> <name_label file_path>
-
<target_folder_path> : Select configure file path. (Generate and save to this path.)
-
<target_data_folder_path> : Select your dataset folder path. (Include text and image file.)
-
<name_label_file_path> : Select your label text file for dataset. (e.g. car , person , ..)
-
Automatic download https://github.com/demulab/divide_files.git and rename .divide_files
License
MIT : https://github.com/Ar-Ray-code/generate_cfg_for_yolo/blob/main/LICENSE