David

So slow...

3

I don't know why the prediction speed of the model is so slow when running the tta?

What is the POSITIVE_KolektorSDD？

2

Hello, can you tell me what's the meaning for POSITIVE_KolektorSDD in config.py? POSITIVE_KolektorSDD=[['5'], ['6'], ['2'], ['3'], ['5'], ['7'], ['1'], ['2'], ['6'], ['3'], ['4'], ['5'], ['3'], ['7'], ['3'], ['5'], ['5'], ['3'],...

Where is the code?

4

Hello, could u release your source code for further research?

yolov5 6.0版本官方yolov5s.pt转换过程

2

 ` onnx_path = "/workspace/custom/detection/yolov5/yolov5s-sim.onnx" prototxt_path = "./yolov5s-60-original.prototxt" caffemodel_path = "./yolov5s-60-original.caffemodel" graph = getGraph(onnx_path) convertToCaffe(graph, prototxt_path, caffemodel_path) ` 请问下如何解决？

About implementation？

2

Hi, thank u for your awesome job! Could u please tell me how to apply this technique to the yolov5 project? Or can u provide a demo tutorial? Thank you...

The download speed is so slow...

4

 Is it any solution to address?

### 💡 Your Question As ↑ point. ### Versions _No response_

微调验证阶段报错：IndexError: list index out of range

31

我这边随便下载了一份demo数据集：[balloons](https://universe.roboflow.com/alon-barenboim/balloons-duyrr/dataset/3)，并根据[finetuning.md](https://github.com/AILab-CVC/YOLO-World/blob/master/docs/finetuning.md)进行修改和配置，下面是对应的配置文件： ```python _base_ = ( '../../third_party/mmyolo/configs/yolov8/' 'yolov8_l_syncbn_fast_8xb16-500e_coco.py') custom_imports = dict( imports=['yolo_world'], allow_failed_imports=False) # hyper-parameters num_classes = 3 num_training_classes = 3 max_epochs = 80 # Maximum training epochs close_mosaic_epochs = 10...

如何导出基于EfficientCSPLayerWithTwoConv结构的onnx模型结构？

11

希望能提供一个导出流程和对应的ort_demo示例。

Supported device？

3

As shown in the config file `config_nas.py`, is it only the V100 or t40 are supported here? If i want to use 2 * 2080Ti, how can i change that?