很好的工作，请问自己的训练的模型怎么修改呢？

[henbucuoshanghai]

你好我 根据安卓demo的cpp代码，只扣出调用tnn，测试图片的代码，在本机运行成功。

我自己训练一个模型，只修改为一类，别的默认官方的，请问为何提示无法调用。模型输出不对吗？

[henbucuoshanghai]

E/tnn: GetOutputMat [File /home/TNN-master/source/tnn/core/instance.cc][Line 250] instance dont have the output with name: 413 0 E/tnn: GetOutputMat [File /home/TNN-master/source/tnn/core/instance.cc][Line 250] instance dont have the output with name: 431 0 Segmentation fault (core dumped)

[henbucuoshanghai]

onnx的输出是1252006 因为1类+5=6 故这是对的 请问tnn的输出难道不是如此吗？ layers是做什么的？我网络结构什么的统统没改，请问怎么修改layers？ 他 是做什么的？ 多谢多谢！！

[henbucuoshanghai]

1252006

[cmdbug]

https://github.com/cmdbug/TNN_Demo/issues/1

[henbucuoshanghai]

你好 按照操作修改，无法export出模型 报错维度对不上。。

[henbucuoshanghai]

Fusing layers... Model Summary: 224 layers, 7053910 parameters, 0 gradients Traceback (most recent call last): File "export.py", line 190, in main(opt) File "export.py", line 185, in main run(vars(opt)) File "export.py", line 147, in run y = model(img) # dry runs File "/anaconda3/envs/lili/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl return forward_call(*input, *kwargs) File "yolov5/models/yolo.py", line 125, in forward return self.forward_once(x, profile, visualize) # single-scale inference, train File "/yolov5/models/yolo.py", line 156, in forward_once x = m(x) # run File "/anaconda3/envs/lili/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl return forward_call(input, kwargs) File "/models/yolo.py", line 70, in forward y = torch.cat((xy, wh, y[..., 4:]), -1) RuntimeError: torch.cat(): Sizes of tensors must match except in dimension 4. Got 1 and 3 in dimension 1 (The offending index is 1)

[cmdbug]

使用v3版本试，最新版本未测试。

[henbucuoshanghai]

感谢 但是v5训练的模型 v3能行吗？结构 某些层什么的改变了吗？

[cmdbug]

yolov5的v3版本，就是转换说明使用的版本。v5能不能用v3的就不知道了，只用过v3版本。

[henbucuoshanghai]

Namespace(batch_size=1, img_size=[640, 640], weights='./last.pt') Traceback (most recent call last): File "export.py", line 29, in model = torch.load(opt.weights, map_location=torch.device('cpu'))['model'].float() File "anaconda3/envs/lili/lib/python3.8/site-packages/torch/serialization.py", line 607, in load return _load(opened_zipfile, map_location, pickle_module, **pickle_load_args) File "/anaconda3/envs/lili/lib/python3.8/site-packages/torch/serialization.py", line 882, in _load result = unpickler.load() File "anaconda3/envs/lili/lib/python3.8/site-packages/torch/serialization.py", line 875, in find_class return super().find_class(mod_name, name) ModuleNotFoundError: No module named 'models'

[henbucuoshanghai]

load model error.