search

HaloTrouvaille / YOLO-Multi-Backbones-Attention

Model Compression—YOLOv3 with multi lightweight backbones(ShuffleNetV2 HuaWei GhostNet), attention, prune and quantization
https://github.com/HaloTrouvaille/YOLO-Multi-Backbones-Attention
493 stars 118 forks source link

visdrone训练参数 #1

Closed lilslsls closed 4 years ago

lilslsls commented 4 years ago

您好，您给出的visdrone数据集的训练结果就是您代码中的默认参数的结果吗

HaloTrouvaille commented 4 years ago

您好，我记得我用的是默认参数。但是Visdrone数据集的训练结果我没有训练完，应该会更高一些。