PureHing
PureHing
@saswat0 sorry,Is there a more detailed description? All about anchor boxes are in this [file](https://github.com/PureHing/face-mask-detection-tf2/blob/fd66b0f1a5a71ee3265c3347740d6fbf3ed7000a/components/prior_box.py#L54)
@southmilkyway 你好，可以分享一下你打包的C# dll吗
@DennisLiu1993 信箱是多少呢
> > @southmilkyway 你好，可以分享一下你打包的C# dll吗 > > > > 可以的，但是博主这里有标准的，可能会更适合你 @southmilkyway 可以邮件发我一下吗？非常感谢！[email protected] solved.
@AyushExel batch=2，the same problem.
@AyushExel Linux server 5.4.0-26-generic #30-Ubuntu SMP Mon Apr 20 16:58:30 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux cuda Info :Ultralytics YOLOv8.0.25 🚀 Python-3.10.9 torch-1.13.1+cu117 CUDA:3 (NVIDIA GeForce RTX 3080 Ti, 12045MiB)...
@JustasBart Use yolov5su.yaml, set the imgsz to 960, rect training, how much memory is theoretically required? > @PureHing Can you confirm that you don't have like a huge number of...
@JustasBart Set worker=8, batch=2 can run, but it still feels abnormal.
@JustasBart > 16 or 15 batch = 16 or 15 ，unable to run normally.
@AyushExel crowdhuman dataset, training with 15000 images