search

AlexeyAB / darknet

YOLOv4 / Scaled-YOLOv4 / YOLO - Neural Networks for Object Detection (Windows and Linux version of Darknet )
http://pjreddie.com/darknet/
Other
21.77k stars 7.96k forks source link

About training map #5684

Open beizhengren opened 4 years ago

beizhengren commented 4 years ago

Hi @AlexeyAB

added drawing of chart of average-Loss and accuracy-mAP (-map flag) during training

When using arg -map, the chart will show the real-time map about current training. And I think when training and testing set are under the same distribution, training mAP sholud be less than testing mAP, for testing weights is better than training. Is it right? Thanks!

AlexeyAB commented 4 years ago

Training mAP is always higher than testing mAP.