LeapLabTHU / EfficientTrain

1.5−3.0× lossless training or pre-training speedup. An off-the-shelf, easy-to-implement algorithm for the efficient training of foundation visual backbones.
MIT License
209 stars 9 forks source link

how to use it in yolo training like yolov8? #8

Open Fischer-pixel opened 5 months ago