search

Liuchen1997 / RFAConv

RAFConv: Innovating Spatital Attention and Standard Convolutional Operation
147 stars 19 forks source link

YOLOv8 #16

Open weiyang-in opened 4 months ago

weiyang-in commented 4 months ago

Why is this error occurring."RuntimeError: Given groups=1, weight of size [32, 16, 1, 1], expected input[1, 128, 64, 64] to have 16 channels, but got 128 channels instead"

CV-ZhangXin commented 4 months ago

@weiyang-in I don't know your specific question. I would like to know how you are using RFAConv in YOLOv8. Please let me know your specific details and I'll be happy to assist you.

Wulingtian commented 3 months ago

我不知道你的具体问题。我想知道您如何在 YOLOv8 中使用 RFAConv。请让我知道您的具体详细信息，我很乐意为您提供帮助。 可否提供一下yolov8使用RFAConv的预训练权重？

CV-ZhangXin commented 1 month ago

@Wulingtian 可以的，但是我不知道你需要哪个数据集的预训练权重，是COCO的还是VOC的？

Wulingtian commented 1 month ago

@Wulingtian 可以的，但是我不知道你需要哪个数据集的预训练权重，是COCO的还是VOC的？

coco的预训练模型

CV-ZhangXin commented 4 weeks ago

@Wulingtian 抱歉这一阵事情比较多，你现在还需要吗？

Wulingtian commented 4 weeks ago

@Wulingtian 抱歉这一阵事情比较多，你现在还需要吗？

要的

CV-ZhangXin commented 4 weeks ago

@Wulingtian 需要YOLOV8具体哪个模型的，就YOLOv8n训练COCO的吗？还是YOLOV8n + RFAConv的。

Wulingtian commented 4 weeks ago

@Wulingtian 需要YOLOV8具体哪个模型的，就YOLOv8n训练COCO的吗？还是YOLOV8n + RFAConv的。

YOLOV8n + RFAConv

CV-ZhangXin commented 3 weeks ago

@Wulingtian ok , 不好意思，最近有点忙，加我qq2643103268，我私发给你。