Python ValueError:检查模型目标时出错:传递给模型的Numpy数组列表的大小不是模型预期的大小
我有多重输出Python ValueError:检查模型目标时出错:传递给模型的Numpy数组列表的大小不是模型预期的大小,python,tensorflow,machine-learning,keras,deep-learning,Python,Tensorflow,Machine Learning,Keras,Deep Learning,我有多重输出 out = [Dense(19, name='one', activation='softmax')(out), Dense(19, name='two', activation='softmax')(out), Dense(19, name='three', activation='softmax')(out), Dense(19, name='four', activation='softmax')(out)]
out = [Dense(19, name='one', activation='softmax')(out),
Dense(19, name='two', activation='softmax')(out),
Dense(19, name='three', activation='softmax')(out),
Dense(19, name='four', activation='softmax')(out)]
model.fit(reshape_train_X, y_onehot, batch_size=400, epochs=100, verbose=2,
validation_split=0.2, callbacks=callbacks_list)
这是我的y_onehot格式:
[array([[1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0]],
dtype=uint8), array([[1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0]],dtype=uint8),.....]
我收到了这个错误信息
ValueError: Error when checking model target: the list of Numpy arrays that you are passing to your model is not the size the model expected. Expected to see 4 array(s), but instead got the following list of 5000 arrays: [array([[1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
...
我不知道为什么当y_onehot在数组中有四个列表时会发生此错误
len(YuOnehot):5000
打印(“y_onehot”,y_onehot[0])
打印(“y_onehot”,len(y_onehot[0]))
我试试看。但还是不起作用
谢谢你的帮助。这是一个虚构的例子。注意你的眼睛。您必须将每个分开的输出进行拟合
inp = Input((50))
x = Dense(32)(inp)
x1 = Dense(19, name='one', activation='softmax')(x)
x2 = Dense(19, name='two', activation='softmax')(x)
x3 = Dense(19, name='three', activation='softmax')(x)
x4 = Dense(19, name='four', activation='softmax')(x)
model = Model(inp, [x1,x2,x3,x4])
model.compile('adam', 'categorical_crossentropy')
X = np.random.uniform(0,1, (5000,50))
y1 = np.random.randint(0,2, (5000,19))
y2 = np.random.randint(0,2, (5000,19))
y3 = np.random.randint(0,2, (5000,19))
y4 = np.random.randint(0,2, (5000,19))
model.fit(X, [y1,y2,y3,y4], epochs=10)
您是否尝试将
y\u onehot
传递到model.fit
作为np.array(y\u onehot)
?是的,但仍然收到此错误消息。这很有效!通过拟合每个分离的输出都是正确的!非常感谢你。
y_onehot 4
inp = Input((50))
x = Dense(32)(inp)
x1 = Dense(19, name='one', activation='softmax')(x)
x2 = Dense(19, name='two', activation='softmax')(x)
x3 = Dense(19, name='three', activation='softmax')(x)
x4 = Dense(19, name='four', activation='softmax')(x)
model = Model(inp, [x1,x2,x3,x4])
model.compile('adam', 'categorical_crossentropy')
X = np.random.uniform(0,1, (5000,50))
y1 = np.random.randint(0,2, (5000,19))
y2 = np.random.randint(0,2, (5000,19))
y3 = np.random.randint(0,2, (5000,19))
y4 = np.random.randint(0,2, (5000,19))
model.fit(X, [y1,y2,y3,y4], epochs=10)