Pytorch 属性错误：'；Conv3d&x27；对象没有属性'；填充模式'；

大家好，可能看到这个的人。我正在尝试将c3d模型转换为onnx，以用于一般用途

C3D列车代码取自：

因此，根据README.MD的要求，c3d模型的原始env为torch=0.4.0python=3.5.2

from network import C3D_model
import torch
import torch.onnx

device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
print("Device being used:", device)


def test():

    model = C3D_model.C3D(num_classes=2)
    checkpoint = torch.load('C3D-ucf101_epoch-3.pth.tar',
                        map_location=lambda storage, loc: storage)
    model.load_state_dict(checkpoint['state_dict'])
    torch.save(model, 'c3d_thunder.pkl')
    dummy_input1 = torch.randn(1, 3, 16, 112, 112)
    torch.onnx.export(model, dummy_input1, "C3D.onnx", verbose=True)

if __name__ == "__main__":
    test()

起初，我试图在原始环境下转换此模型，但遇到以下错误：

Traceback (most recent call last):
  File "/workspace/pytorch-video-recognition/model2onnx.py", line 31, in <module>
    test()
  File "/workspace/pytorch-video-recognition/model2onnx.py", line 27, in test
    torch.onnx.export(model, dummy_input1, "C3D.onnx", verbose=True)
  File "/usr/local/lib/python3.5/dist-packages/torch/onnx/__init__.py", line 25, in export
    return utils.export(*args, **kwargs)
  File "/usr/local/lib/python3.5/dist-packages/torch/onnx/utils.py", line 84, in export
    _export(model, args, f, export_params, verbose, training, input_names, output_names)
  File "/usr/local/lib/python3.5/dist-packages/torch/onnx/utils.py", line 154, in _export
    _onnx_opset_version, defer_weight_export)
RuntimeError: ONNX export failed: Couldn't export operator aten::max_pool3d

仍有错误：

Traceback (most recent call last):
  File "C:/Users/google/Desktop/onnx_test/t1.py", line 33, in <module>
    test()
  File "C:/Users/google/Desktop/onnx_test/t1.py", line 29, in test
    torch.onnx.export(model, dummy_input1, "C3D.onnx", verbose=True)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\onnx\__init__.py", line 25, in export
    return utils.export(*args, **kwargs)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\onnx\utils.py", line 131, in export
    strip_doc_string=strip_doc_string)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\onnx\utils.py", line 363, in _export
    _retain_param_name, do_constant_folding)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\onnx\utils.py", line 266, in _model_to_graph
    graph, torch_out = _trace_and_get_graph_from_model(model, args, training)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\onnx\utils.py", line 225, in _trace_and_get_graph_from_model
    trace, torch_out = torch.jit.get_trace_graph(model, args, _force_outplace=True)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\jit\__init__.py", line 231, in get_trace_graph
    return LegacyTracedModule(f, _force_outplace, return_inputs)(*args, **kwargs)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\nn\modules\module.py", line 493, in __call__
    result = self.forward(*input, **kwargs)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\jit\__init__.py", line 294, in forward
    out = self.inner(*trace_inputs)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\nn\modules\module.py", line 491, in __call__
    result = self._slow_forward(*input, **kwargs)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\nn\modules\module.py", line 481, in _slow_forward
    result = self.forward(*input, **kwargs)
  File "C:\Users\google\Desktop\onnx_test\network\C3D_model.py", line 46, in forward
    x = self.relu(self.conv1(x))
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\nn\modules\module.py", line 491, in __call__
    result = self._slow_forward(*input, **kwargs)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\nn\modules\module.py", line 481, in _slow_forward
    result = self.forward(*input, **kwargs)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\nn\modules\conv.py", line 468, in forward
    if self.padding_mode == 'circular':
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\nn\modules\module.py", line 539, in __getattr__
    type(self).__name__, name))
AttributeError: 'Conv3d' object has no attribute 'padding_mode'

回溯（最近一次呼叫最后一次）：
文件“C:/Users/google/Desktop/onnx_test/t1.py”，第33行，在
测试（）
文件“C:/Users/google/Desktop/onnx_test/t1.py”，第29行，在测试中
torch.onnx.export（模型，虚拟输入1，“C3D.onnx”，verbose=True）
文件“D:\Anaconda3\envs\torch\u learn\lib\site packages\torch\onnx\\uuuuuu init\uuuuuu.py”，第25行，在导出中
返回utils.export（*args，**kwargs）
导出中的文件“D:\Anaconda3\envs\torch\u learn\lib\site packages\torch\onnx\utils.py”，第131行
条带文档字符串=条带文档字符串）
文件“D:\Anaconda3\envs\torch\u learn\lib\site packages\torch\onnx\utils.py”，第363行，在导出中
_保留参数名称，执行常数折叠）
文件“D:\Anaconda3\envs\torch\u learn\lib\site packages\torch\onnx\utils.py”，第266行，在模型图中
graph，torch_out=_trace_和_get_graph_from_model（model，args，training）
文件“D:\Anaconda3\envs\torch\u learn\lib\site packages\torch\onnx\utils.py”，第225行，在跟踪和从模型获取图形中
trace，torch\u out=torch.jit.get\u trace\u图（model，args，\u force\u outplace=True）
文件“D:\Anaconda3\envs\torch\u learn\lib\site packages\torch\jit\uuuu init\uuuu.py”，第231行，在get\u trace\u图中
返回LegacyTracedModule（f，力输出，返回输入）（*args，**kwargs）
文件“D:\Anaconda3\envs\torch\u learn\lib\site packages\torch\nn\modules\module.py”，第493行，在调用中__
结果=自我转发（*输入，**kwargs）
文件“D:\Anaconda3\envs\torch\u learn\lib\site packages\torch\jit\uuuu init\uuuu.py”，第294行，向前
输出=内部（*跟踪输入）
文件“D:\Anaconda3\envs\torch\u learn\lib\site packages\torch\nn\modules\module.py”，第491行，在调用中__
结果=self.\u slow\u forward（*输入，**kwargs）
文件“D:\Anaconda3\envs\torch\u learn\lib\site packages\torch\nn\modules\module.py”，第481行，慢速前进
结果=自我转发（*输入，**kwargs）
文件“C:\Users\google\Desktop\onnx\u test\network\C3D\u model.py”，第46行，向前
x=self.relu（self.conv1（x））
文件“D:\Anaconda3\envs\torch\u learn\lib\site packages\torch\nn\modules\module.py”，第491行，在调用中__
结果=self.\u slow\u forward（*输入，**kwargs）
文件“D:\Anaconda3\envs\torch\u learn\lib\site packages\torch\nn\modules\module.py”，第481行，慢速前进
结果=自我转发（*输入，**kwargs）
文件“D:\Anaconda3\envs\torch\u learn\lib\site packages\torch\nn\modules\conv.py”，第468行，向前
如果self.padding_mode==“循环”：
文件“D:\Anaconda3\envs\torch\u learn\lib\site packages\torch\nn\modules\module.py”，第539行，位于__
类型（自我）。\uuuuu名称（自我，名称））
AttributeError:“Conv3d”对象没有“padding\u mode”属性

---

有人能告诉我如何解决这个问题吗？谢谢你的时间~

您是否尝试过使用最新版本的PyTorch

您是否尝试过使用最新版本的PyTorch

你应该发布你的代码，不要复制图像，这样更容易理解。很抱歉，我已经更新了代码和错误内容。你应该发布代码，不要复制图像，这样更容易理解。很抱歉，我已经更新了代码和错误内容。

Traceback (most recent call last):
  File "C:/Users/google/Desktop/onnx_test/t1.py", line 33, in <module>
    test()
  File "C:/Users/google/Desktop/onnx_test/t1.py", line 29, in test
    torch.onnx.export(model, dummy_input1, "C3D.onnx", verbose=True)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\onnx\__init__.py", line 25, in export
    return utils.export(*args, **kwargs)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\onnx\utils.py", line 131, in export
    strip_doc_string=strip_doc_string)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\onnx\utils.py", line 363, in _export
    _retain_param_name, do_constant_folding)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\onnx\utils.py", line 266, in _model_to_graph
    graph, torch_out = _trace_and_get_graph_from_model(model, args, training)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\onnx\utils.py", line 225, in _trace_and_get_graph_from_model
    trace, torch_out = torch.jit.get_trace_graph(model, args, _force_outplace=True)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\jit\__init__.py", line 231, in get_trace_graph
    return LegacyTracedModule(f, _force_outplace, return_inputs)(*args, **kwargs)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\nn\modules\module.py", line 493, in __call__
    result = self.forward(*input, **kwargs)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\jit\__init__.py", line 294, in forward
    out = self.inner(*trace_inputs)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\nn\modules\module.py", line 491, in __call__
    result = self._slow_forward(*input, **kwargs)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\nn\modules\module.py", line 481, in _slow_forward
    result = self.forward(*input, **kwargs)
  File "C:\Users\google\Desktop\onnx_test\network\C3D_model.py", line 46, in forward
    x = self.relu(self.conv1(x))
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\nn\modules\module.py", line 491, in __call__
    result = self._slow_forward(*input, **kwargs)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\nn\modules\module.py", line 481, in _slow_forward
    result = self.forward(*input, **kwargs)
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\nn\modules\conv.py", line 468, in forward
    if self.padding_mode == 'circular':
  File "D:\Anaconda3\envs\torch_learn\lib\site-packages\torch\nn\modules\module.py", line 539, in __getattr__
    type(self).__name__, name))
AttributeError: 'Conv3d' object has no attribute 'padding_mode'