关于模型的用法和数据集

[LoveYstT]

1. 项目中有yolov10n_backbone.yaml和yolov10n_lightneck.yaml两个模型，在train.py里model = YOLO(' ')使用应该哪个呀？
2. SCoralDet数据集一共有多少个nc？分别都是什么？有的class_id是-1，这是为什么呀？

[RDXiaoLu]

1. The yolov10n_backbone.yaml corresponds to improvements in the MPFB, while the lightneck.yaml corresponds to improvements in the neck. You can combine them together as needed.

2. The SCoralDet dataset contains 6 types of coral, with class_id ranging from 0 to 5. There should not be any class_id with a value of -1.

Thank you for your interest in my work.

[LoveYstT]

1. 可以详细说一下怎么结合吗，我是初学实在不太会，麻烦了
2. 比如这第4个的class_id就是-1
   

[RDXiaoLu]

1. I’ve combined them for you, and you can check the file at lib/cfg/models/v10/yolov10n_combine.yaml.

2. Is this the annotation file generated by running coco2yolo.py? There might have been an error in the earlier steps. I recommend carefully checking whether a -1 appeared during the previous steps.

[LoveYstT]

1. 非常感谢
2. 好的，我再检查一下前面的步骤

[LoveYstT]

task.py里的parse_model函数是怎么修改的呀

[RDXiaoLu]

For example:

for i, (f, n, m, args) in enumerate(d["backbone"] + d["head"]):  # from, number, module, args
        try:
            if m == 'node_mode':
                m = d[m]
                if len(args) > 0:
                    if args[0] == 'head_channel':
                        args[0] = int(d[args[0]])
            t = m
            m = getattr(torch.nn, m[3:]) if 'nn.' in m else globals()[m]  # get module
        except:
            pass
        for j, a in enumerate(args):
            if isinstance(a, str):
                with contextlib.suppress(ValueError):
                    try:
                        args[j] = locals()[a] if a in locals() else ast.literal_eval(a)
                    except:
                        args[j] = a
        n = n_ = max(round(n * depth), 1) if n > 1 else n  # depth gain
        if m in {
            C2f_MPFB
        }:
        if args[0] == 'head_channel':
                args[0] = d[args[0]]
            c1, c2 = ch[f], args[0]
            if c2 != nc:  # if c2 not equal to number of classes (i.e. for Classify() output)
                c2 = make_divisible(min(c2, max_channels) * width, 8)
           
            args = [c1, c2, *args[1:]]
            if m in {
                     C2f_MPFB
                     }:
                args.insert(2, n)  # number of repeats
                n = 1

[LoveYstT]

感谢

[RDXiaoLu]

Thank you for bringing this up. If you find the project useful, a ⭐ on the repository would be greatly appreciated.

[LoveYstT]

好的，作为初学者也谢谢你的帮助

[LoveYstT]

在task.py里的parse_model函数有添加VoVGSCSP相关的代码嘛

[RDXiaoLu]

Just like adding C2f_MPFB, you only need to:

if m in {
            C2f_MPFB, VoVGSCSP 
        }:

[LoveYstT]

应该加到哪个if里面，你上面发的那段代码有两个
if m in {
C2f_MPFB
}: