运行时错误:mat1 和 mat2 形状无法相乘(200x16 和 32x1)
我想连接两个全连接层。然后,在连接它们之后,我想构建另一个具有全连接层的神经网络。
我可以看到该错误是由于未正确设置 cat_x = torch.cat([x, x1])
引起的。但是,我不知道如何解决这个问题。
import torch
from torch import nn, optim
import numpy as np
from matplotlib import pyplot as plt
class Regression(nn.Module):
def __init__(self):
super().__init__()
self.linear1 = nn.Linear(2, 32)
self.linear2 = nn.Linear(32, 16)
self.linear3 = nn.Linear(32, 1)
def forward(self, input):
x = nn.functional.relu(self.linear1(input))
x = nn.functional.relu(self.linear2(x))
x1 = nn.functional.elu(self.linear1(input))
x1 = nn.functional.elu(self.linear2(x1))
cat_x = torch.cat([x, x1])
cat_x = self.linear3(cat_x)
return cat_x
def train(model, optimizer, E, iteration, x, y):
losses = []
for i in range(iteration):
optimizer.zero_grad() # 勾配情報を0に初期化
y_pred = model(x) # 予測
loss = E(y_pred.reshape(y.shape), y) # 損失を計算(shapeを揃える)
loss.backward() # 勾配の計算
optimizer.step() # 勾配の更新
losses.append(loss.item()) # 損失値の蓄積
print('epoch=', i+1, 'loss=', loss)
return model, losses
def test(model, x):
y_pred = model(x).data.numpy().T[0] # 予測
return y_pred
x = np.random.uniform(0, 10, 100) # x軸をランダムで作成
y = np.random.uniform(0.9, 1.1, 100) * np.sin(2 * np.pi * 0.1 * x) # 正弦波を作成
x = torch.from_numpy(x.astype(np.float32)).float() # xをテンソルに変換
y = torch.from_numpy(y.astype(np.float32)).float() # yをテンソルに変換
X = torch.stack([torch.ones(100), x], 1) # xに切片用の定数1配列を結合
net = Regression()
optimizer = optim.RMSprop(net.parameters(), lr=0.01) # 最適化にRMSpropを設定
E = nn.MSELoss()
net, losses = train(model=net, optimizer=optimizer, E=E, iteration=5000, x=X, y=y)
错误信息
/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py in linear(input, weight, bias)
1846 if has_torch_function_variadic(input, weight, bias):
1847 return handle_torch_function(linear, (input, weight, bias), input, weight, bias=bias)
-> 1848 return torch._C._nn.linear(input, weight, bias)
1849
1850
RuntimeError: mat1 and mat2 shapes cannot be multiplied (200x16 and 32x1)
I would like to connect two Fully Connected layers. Then, after connecting them, I want to build another neural net with the Fully Connected layer.
I can see that the error is caused by not setting cat_x = torch.cat([x, x1])
properly. However, I do not know how to solve this problem.
import torch
from torch import nn, optim
import numpy as np
from matplotlib import pyplot as plt
class Regression(nn.Module):
def __init__(self):
super().__init__()
self.linear1 = nn.Linear(2, 32)
self.linear2 = nn.Linear(32, 16)
self.linear3 = nn.Linear(32, 1)
def forward(self, input):
x = nn.functional.relu(self.linear1(input))
x = nn.functional.relu(self.linear2(x))
x1 = nn.functional.elu(self.linear1(input))
x1 = nn.functional.elu(self.linear2(x1))
cat_x = torch.cat([x, x1])
cat_x = self.linear3(cat_x)
return cat_x
def train(model, optimizer, E, iteration, x, y):
losses = []
for i in range(iteration):
optimizer.zero_grad() # 勾配情報を0に初期化
y_pred = model(x) # 予測
loss = E(y_pred.reshape(y.shape), y) # 損失を計算(shapeを揃える)
loss.backward() # 勾配の計算
optimizer.step() # 勾配の更新
losses.append(loss.item()) # 損失値の蓄積
print('epoch=', i+1, 'loss=', loss)
return model, losses
def test(model, x):
y_pred = model(x).data.numpy().T[0] # 予測
return y_pred
x = np.random.uniform(0, 10, 100) # x軸をランダムで作成
y = np.random.uniform(0.9, 1.1, 100) * np.sin(2 * np.pi * 0.1 * x) # 正弦波を作成
x = torch.from_numpy(x.astype(np.float32)).float() # xをテンソルに変換
y = torch.from_numpy(y.astype(np.float32)).float() # yをテンソルに変換
X = torch.stack([torch.ones(100), x], 1) # xに切片用の定数1配列を結合
net = Regression()
optimizer = optim.RMSprop(net.parameters(), lr=0.01) # 最適化にRMSpropを設定
E = nn.MSELoss()
net, losses = train(model=net, optimizer=optimizer, E=E, iteration=5000, x=X, y=y)
error message
/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py in linear(input, weight, bias)
1846 if has_torch_function_variadic(input, weight, bias):
1847 return handle_torch_function(linear, (input, weight, bias), input, weight, bias=bias)
-> 1848 return torch._C._nn.linear(input, weight, bias)
1849
1850
RuntimeError: mat1 and mat2 shapes cannot be multiplied (200x16 and 32x1)
如果你对这篇内容有疑问,欢迎到本站社区发帖提问 参与讨论,获取更多帮助,或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱,如果其他用户或者作者回复了您的评论,将不能在第一时间通知您!
发布评论
评论(2)
由于
x
和x1
的维度均为(100,16)
,因此torch.cat
运算符在第一维度(因为它们在该方向上的大小相似)。为了使您的代码正常工作,请将cat_x = torch.cat([x, x1])
更改为cat_x = torch.cat([x, x1], dim=1)
since both
x
andx1
have dimensions of(100,16)
, thetorch.cat
operator concatenates in the 1st dimension (since they are of similar size in that direction). For your code to work change thecat_x = torch.cat([x, x1])
tocat_x = torch.cat([x, x1], dim=1)
尝试:
Try: