-
1、我想要完成一个训练数据集较少的任务，大概2000-3000张。但是现在yolov10n网络出现了严重的过拟合问题，在train上达到了map:.5/map:.95：0.98/0.82的精度，但是在test集上表现很差,map:.5只有0.4。我想问下除了dropout参数可以避免过拟合，还有什么办法吗？
2、另外，我看到如果使用nms会在一些小型网络如YOLOv10n上有一定的精度提升，是否…
-
I have implemented the onnx to tensorrt for yolov10m.
Dynamic batching is not supported.
In onnx, we have the following output.
```text
name: output0
tensor: float32[1,300,6]
```
300 is t…
-
YOLOv10 has an input argument iou which is by default 0.7. Please elaborate, What is meant by iou arguments and what is it's affect?
Thanks!
-
First of all, thank you for your excellent work. I'm doing mobile model app development，I would like to ask how to convert the Yolov10 model to the NCNN infer. Can you provide some detailed steps？tha…
-
### 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 fou…
-
Creating and training YOLO model
Traceback (most recent call last):
File "/home/vrai/train_server/core.py", line 183, in
main('settings.conf')
File "/home/vrai/train_server/core.py", line…
-
(labelImgEnv) C:\Meet\new>yolo task=detect mode=predict model=best.pt show=True conf=0.5 source=w.jpg
Ultralytics YOLOv8.1.34 🚀 Python-3.12.3 torch-2.3.0+cu121 CUDA:0 (NVIDIA GeForce GTX 1650 Ti, 409…
-
1. yolov10 导出onnx，output是[1,300,6]，6这个维度明显不对，coco 80类怎么说也不可能是6维，rt-detr导出是[1,300,84]，维度是对的。v1.1和master都测试了，都有这问题
2. yolov10自己的yolov10s.pt可以正常推理，但自己训练的pt在predict时候说dict没有shape属性，错在opy.py里非极大值抑制里。v1.1和…
-
Great project !!!
How can i infer a dataset in batch and multi-GPU?
-
### Search before asking
- [X] I have searched the Yolo Tracking [issues](https://github.com/mikel-brostrom/yolo_tracking/issues) and found no similar bug report.
### Question
Hey, thanks f…