Hu Kai

v5Lite-e

1

请问为什么v5Lite-e中的head用的是DWConv替换C3呢，而不继续像backbone那样用Shuffle_Block替换呢？ 

YOLOLayer的forward的变化

1

 为什么reference的时候要改写成这样呢？

I have a doubt

1

Is this line suppose to be TPs? 

Unet预训练权重

1

你好，这个unet仓库没有大型数据集的预训练权重的吗？

3.0一直报错

17

Traceback (most recent call last): File "/home/yyds/hukai/faster-rcnn-pytorch-master/train.py", line 382, in fit_one_epoch(model, train_util, loss_history, optimizer, epoch, epoch_step, epoch_step_val, gen, gen_val, UnFreeze_Epoch, Cuda, fp16, scaler, save_period, save_dir) File "/home/yyds/hukai/faster-rcnn-pytorch-master/utils/utils_fit.py", line 27, in...

yolov8n's coco pre-training?

2

### Search before asking - [X] I have searched the YOLOv8 [issues](https://github.com/ultralytics/ultralytics/issues) and [discussions](https://github.com/ultralytics/ultralytics/discussions) and found no similar questions. ### Question Can you send me a copy of the training...

num_features ??? 

Do you have CSPDenseNet pre_train weight?

有个问题：在内存复用时，如果当前需要的内存小于已申请的内存，直接返回已申请的内存，但是如何保证这段内存没有在使用，如果在其他线程中这段内存正在被使用，那不就错了？

想实现的功能：使用mmp代替cv::Mat image = cv::imdecode(cv::Mat(1, buf.bytesused, CV_8UC1, buffers_[buf.index].start), cv::IMREAD_COLOR); 减少cpu使用率 代码： #include #include #include #include #include #include #include #include #include #include #include // #include #include #include #include #define HEIGHT 480 #define...