8
1

Delete article

Deleted articles cannot be recovered.

Draft of this article would be also deleted.

Are you sure you want to delete this article?

More than 3 years have passed since last update.

pytorchでcannot assign module before Module.init() callが出た時

Posted at

すごく簡単な話なのですが、いつかやらかして長時間溶かしそうなので備忘録です。

##該当コードとエラー内容

import torch
import torch.nn as nn

class Encoder(nn.Module):
    def __init__(self, p_n_features_num, timesteps) -> None:
        self.p_n_features_num = p_n_features_num
        self.linear = nn.Linear(
            p_n_features_num, p_n_features_num, bias=True)
        self.leakyrelu = nn.LeakyReLU()

    def forward(self, input_net):
        input_net = input_net.view(input_net.size(0), self.p_n_features_num)
        return self.leakyrelu(self.linear(input_net))

ここでEncoderの__init__を呼び出すと

cannot assign module before Module.init() call
```のエラーが出ます。

##解決策、修正コード
これはinitのときにnn.Moduleを継承しているのでsuperメソッドを始めに呼び出さなかったのが原因です。そのため、以下のように修正してあげれば直ります。

```python
import torch
import torch.nn as nn

class Encoder(nn.Module):
    def __init__(self, p_n_features_num, timesteps) -> None:
        super(Encoder, self).__init__()
        self.p_n_features_num = p_n_features_num
        self.linear = nn.Linear(
            p_n_features_num, p_n_features_num, bias=True)
        self.leakyrelu = nn.LeakyReLU()

    def forward(self, input_net):
        input_net = input_net.view(input_net.size(0), self.p_n_features_num)
        return self.leakyrelu(self.linear(input_net))
8
1
0

Register as a new user and use Qiita more conveniently

  1. You get articles that match your needs
  2. You can efficiently read back useful information
  3. You can use dark theme
What you can do with signing up
8
1

Delete article

Deleted articles cannot be recovered.

Draft of this article would be also deleted.

Are you sure you want to delete this article?