WeiDu

> 我认为那一条线是存在的，不过位置不对，应该在split下面 你看下面的代码 > > ```python > def forward(self, x): > """Forward pass through C2f layer.""" > y = list(self.cv1(x).chunk(2, 1)) # 这一行的结果y包含了split的2个结果，2个分支都被保存 > y.extend(m(y[-1]) for m in self.m) >...

> 我看了一下最新的图 https://github.com/open-mmlab/mmyolo/tree/main/configs/yolov8 ，发现相比你发的图片删除了一个分支，就是最下面的DarknetBottleneck左侧的分支被删除了，删除了这个分支再加上我上面说的红色分支，最后的concat结果就是对的了。 > > 下图中蓝圈内的分支在新的图片中已经删除了 好的，十分感谢

I tried to change learninga rate .but it didn't work

The dataset was made from a few images I randomly selected from coco2017

You can try to install mmdet 2.28.2 with pip：pip install mmdet == 2.28.2 

 maybe you have a version mismatch problem and you should make sure their versions meet the following requirements 

> Can you share your dataflow description yaml? of course, my codes in dataflow description yaml are as follows: ```yaml nodes: - id: webcam_D435i operator: python: ../webcam_D435i.py inputs: tick: dora/timer/millis/50...

I want to receive four data streams and visualize them at the same time.When I received four types of data for visualization testing one by one, I found that color...

> I was not able to reproduce your error with a basic webcam operator. > > My dataflow > > ```yaml > nodes: > - id: webcam > operator: >...