search

AlexeyAB / darknet

YOLOv4 / Scaled-YOLOv4 / YOLO - Neural Networks for Object Detection (Windows and Linux version of Darknet )
http://pjreddie.com/darknet/
Other
21.65k stars 7.96k forks source link

python API for training #6621

Open r-isr opened 4 years ago

r-isr commented 4 years ago

Hi, Could you please add python API for training as well?

arnaud-nt2i commented 3 years ago

hi @r-isr within commit 9ecc4f7d6b24dec4d56a1d5c480fe842625dc57b dontcare committed on 8 Aug You can work from this: ` import os import sys

os.environ['DARKNET_PATH'] = '/usr/local/darknet' sys.path.append(os.environ['DARKNET_PATH'])

import darknet

def main(): network, class_names, class_colors = darknet.load_network( '../nscan-train/data/cfg/yolov4-tiny-3l-numbers-12.cfg', '../nscan-train/data/numbers.data', './data/weights/numbers/yolov4-tiny-3l-numbers-12_last.weights', batch_size=1 )

if name == 'main': main() `

fabian57fabian commented 3 years ago

hi @r-isr within commit 9ecc4f7 dontcare committed on 8 Aug

os.environ['DARKNET_PATH'] = '/usr/local/darknet' sys.path.append(os.environ['DARKNET_PATH'])

This saved me. Just add those two lines and you can call darknet without getting this error: OSError: ./libdarknet.so: cannot open shared object file: No such file or directory

Also i would suggest suggest, if calling darknet.py script from another in different folder (using import darknet) and getting this error: ModuleNotFoundError: No module named 'darknet'

Just add sys.path.insert(0, os.path.dirname(os.path.abspath(file))) before calling that python script.