Question

（回复bug：https://github.com/zihualiu/pytorch_linear_bug）

我最近在Pytorch遇到了一个奇怪的错误，我希望你能帮助我。在我的一个网络中，我有一个完全连接的层，表示为net.fc_h1。然而，在训练期间，我意识到该层在激活之前输出NaN。所以我把它放在pdb中，希望它会让我产生一些东西。以下是日志：

# in network declaration:
def forward(self, obs):
    z1 = self.fc_h1(obs)
    if np.isnan(np.sum(z1.data.numpy())):
        pdb.set_trace()
    h1 = F.tanh(z1)
    ...

确实捕获了NaN，但我在pdb中意识到如果再次运行该操作，结果将是显着的：

(Pdb) z1.sum()
Variable containing:
nan
[torch.FloatTensor of size 1]

(Pdb) self.fc_h1(obs).sum()
Variable containing:
771.5120
[torch.FloatTensor of size 1]

当我检查我的输入或重量是否包含NaN时，我得到以下内容：（Pdb）self.fc_h1.weight.max（）变量包含： 0.2482 [大小为1的torch.FloatTensor]

(Pdb) self.fc_h1.weight.mean()
Variable containing:
1.00000e-03 *
  1.7761
[torch.FloatTensor of size 1]

(Pdb) self.fc_h1.weight.min()
Variable containing:
-0.2504
[torch.FloatTensor of size 1]

(Pdb) obs.max()
Variable containing:
 6.9884
[torch.FloatTensor of size 1]

(Pdb) obs.min()
Variable containing:
-6.7855
[torch.FloatTensor of size 1]

(Pdb) obs.mean()
Variable containing:
1.00000e-02 *
 -1.5033
[torch.FloatTensor of size 1] 
(Pdb) self.fc_h1.bias.max()
Variable containing:
 0.2482
[torch.FloatTensor of size 1]

(Pdb) self.fc_h1.bias.mean()
Variable containing:
1.00000e-03 *
  3.9104
[torch.FloatTensor of size 1]

(Pdb) self.fc_h1.bias.min()
Variable containing:
-0.2466
[torch.FloatTensor of size 1]

似乎输入，重量和偏见都很好。如果一切都形成良好，线性层如何产生NaN的任何见解？

编辑：更奇怪所以我试图再次进行前锋传球，有趣的是，多次前锋传球给了我不同的结果：

(Pdb) self.fc_h1(obs)
Variable containing:
 2.2321e-01 -6.2586e-01 -1.9004e-01  ...  -4.2521e-01  8.6175e-01  8.6866e-01
-7.2699e-02  7.8234e-01 -5.8862e-01  ...   2.4041e-01 -1.7577e-01  6.9928e-01
-7.2699e-02  7.8234e-01 -5.8862e-01  ...   2.4041e-01 -1.7577e-01  6.9928e-01
                ...                   ⋱                   ...
-6.4686e-02 -1.5819e+00  5.7410e-01  ...  -6.4127e-01  5.2837e-01 -1.3166e+00
 3.9214e-01  2.8727e-01 -5.5699e-01  ...  -8.3164e-01 -5.1795e-01 -3.7637e-01
-9.6061e-01  1.4780e-01  5.3614e-02  ...  -1.5042e+00  6.0759e-02 -3.6862e-01
[torch.FloatTensor of size 4096x170]

(Pdb) self.fc_h1(obs)
Variable containing:
 2.2321e-01 -6.2586e-01 -1.9004e-01  ...  -4.2521e-01  8.6175e-01  8.6866e-01
-7.2699e-02  7.8234e-01 -5.8862e-01  ...   2.4041e-01 -1.7577e-01  6.9928e-01
-7.2699e-02  7.8234e-01 -5.8862e-01  ...   2.4041e-01 -1.7577e-01  6.9928e-01
                ...                   ⋱                   ...
        nan         nan         nan  ...          nan  5.2837e-01 -1.3166e+00
        nan         nan         nan  ...          nan -5.1795e-01 -3.7637e-01
        nan         nan         nan  ...          nan  6.0759e-02 -3.6862e-01
[torch.FloatTensor of size 4096x170]

我也没有使用GPU，只是CPU。

Answer 1

对我来说，我正在复制RNN名称分类示例中的代码。我添加了优化器和标准模式，而该示例是手动操作并手动更新权重。我无意中向优化器添加了动量值，这就是导致我遇到问题的原因。将动量设置为def remove(self): '''------------------------------------------------------- Removes and returns value from the queue. Use: v = cq.remove() ------------------------------------------------------- Returns: value - the value at the front of the queue - the value is removed from the queue (?) -------------------------------------------------------''' assert (self._count > 0), 'Cannot remove from an empty queue' value = self._values[self._front] self._front = (self._front + 1) % self._count self._count += -1 return value的默认值可修复该问题。

PyTorch nn.Linear layer输出良好的输入和权重

1 个答案: