替换ReID模块

[LiuYuan84]

问题确认 Search before asking

• 我已经搜索过问题，但是没有找到解答。I have searched the question and found no related answer.

请提出你的问题 Please ask your question

您好 我想用VerReID替换掉现有的ReID模块 除了替换模型和路径之外还需要做什么改动吗？ 比如输出路径这种 期待您的回复！！

[zhangyubo0722]

对于模型结构是有变动吗

[LiuYuan84]

不好意思我不太懂这个模型结构是啥意思😢 但是我今天尝试格式转换（onnx -> paddle）的时候总是不成功
C:\Users\AGRSAdmin.conda\envs\ppd\lib\site-packages\paddle\nn\layer\layers.py:2084: UserWarning: Skip loading for embedding0.weight. embedding0.weight receives a shape [5, 2, 768], but the expected shape is [5, 2].
warnings.warn(f"Skip loading for {key}. " + str(err))
C:\Users\AGRSAdmin.conda\envs\ppd\lib\site-packages\paddle\jit\dy2static\program_translator.py:712: UserWarning: full_graph=False don't support input_spec arguments. It will not produce any effect.
You can set full_graph=True, then you can assign input spec.

warnings.warn(
In transformed code:

File "E:\Yuan\project\PaddleDetection\output_inference/VersReID\x2paddle_code.py", line 439, in forward
    x2paddle_191 = paddle.concat(x=[x2paddle_190, x2paddle_175], axis=1)
    x2paddle_192 = paddle.add(x=x2paddle_191, y=x2paddle_base_pos_embed)
    x2paddle_198 = paddle.concat(x=[x2paddle_192, x2paddle_197], axis=1)
    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ <--- HERE
    x2paddle_209 = self.x2paddle_209(x2paddle_198)
    x2paddle_210 = paddle.shape(input=x2paddle_209)

File "C:\Users\AGRSAdmin\.conda\envs\ppd\lib\site-packages\paddle\tensor\manipulation.py", line 1332, in concat
    helper.append_op(
File "C:\Users\AGRSAdmin\.conda\envs\ppd\lib\site-packages\paddle\base\layer_helper.py", line 44, in append_op
    return self.main_program.current_block().append_op(*args, **kwargs)
File "C:\Users\AGRSAdmin\.conda\envs\ppd\lib\site-packages\paddle\base\framework.py", line 4467, in append_op
    op = Operator(
File "C:\Users\AGRSAdmin\.conda\envs\ppd\lib\site-packages\paddle\base\framework.py", line 3214, in __init__
    self.desc.infer_shape(self.block.desc)

ValueError: (InvalidArgument) The 2-th dimension of input[0] and input[1] is expected to be equal.But received input[0]'s shape = [-1, 193, 768], input[1]'s shape = [1, 1, 2].

[Hint: Expected inputs_dims[0][j] == inputs_dims[i][j], but received inputs_dims[0][j]:768 != inputs_dims[i][j]:2.] (at ..\paddle/phi/kernels/funcs/concat_funcs.h:83)
[operator < concat > error]

[zhangyubo0722]

转换的命令是什么呢