Keras 中的训练变分自动编码器引发"InvalidArgumentError: Incompatible shapes"错误

Question

我整个晚上都在尝试让这个VAE工作，但一次又一次地遇到同样的问题。我不确定问题出在哪里。我试过删除回调、验证、更改损失函数、更改采样方法。错误(下面显示的是提前停止(一直是添加到fit函数的最后一个参数。我不知道如何让它发挥作用。

下面是可复制的代码，然后是我一直存在的错误。请注意，更改批次大小确实会更改错误，但不匹配的数字也会随着批次大小而减少。

import pandas as pd
from sklearn.datasets import make_blobs 
from sklearn.preprocessing import MinMaxScaler
import keras.backend as K
import tensorflow as tf
from keras.layers import Input, Dense, Lambda, Layer, Add, Multiply
from keras.models import Model, Sequential
from keras.callbacks import EarlyStopping, LearningRateScheduler
from keras.objectives import binary_crossentropy

x, labels = make_blobs(n_samples=150000, n_features=110,  centers=16, cluster_std=4.0)
scaler = MinMaxScaler()
x = scaler.fit_transform(x)
x = pd.DataFrame(x)
train = x.sample(n = 100000)
train_indexs = train.index.values
test = x[~x.index.isin(train_indexs)]
print(train.shape, test.shape)
min_dim = 2
batch_size = 1024
def sampling(args):
mu, log_sigma = args
eps = K.random_normal(shape=(batch_size, min_dim), mean = 0.0, stddev = 1.0)
return mu + K.exp(0.5 * log_sigma) * eps
#Encoder
inputs = Input(shape=(x.shape[1],))
down1 = Dense(64, activation='relu')(inputs)
mu = Dense(min_dim, activation='linear')(down1)
log_sigma = Dense(min_dim, activation='linear')(down1)
#Sampling
sample_set = Lambda(sampling, output_shape=(min_dim,))([mu, log_sigma])
#decoder
up1 = Dense(64, activation='relu')(sample_set)
output = Dense(x.shape[1], activation='sigmoid')(up1)
vae = Model(inputs, output)
encoder = Model(inputs, mu)
def vae_loss(y_true, y_pred):
recon  = binary_crossentropy(y_true, y_pred)
kl = - 0.5 * K.mean(1 + log_sigma - K.square(mu) - K.exp(log_sigma), axis=-1)
return recon + kl
vae.compile(optimizer='adam', loss=vae_loss)
vae.fit(train, train, shuffle = True, epochs = 1000, 
batch_size = batch_size, validation_data = (test, test), 
callbacks = [EarlyStopping(patience=50)])

错误：

File "<ipython-input-2-7aa4be21434d>", line 62, in <module>
callbacks = [EarlyStopping(patience=50)])
File "C:Usersse01040434Anaconda3libsite-packageskerasenginetraining.py", line 1239, in fit
validation_freq=validation_freq)
File "C:Usersse01040434Anaconda3libsite-packageskerasenginetraining_arrays.py", line 196, in fit_loop
outs = fit_function(ins_batch)
File "C:Usersse01040434Anaconda3libsite-packagestensorflowpythonkerasbackend.py", line 3792, in __call__
outputs = self._graph_fn(*converted_inputs)
File "C:Usersse01040434Anaconda3libsite-packagestensorflowpythoneagerfunction.py", line 1605, in __call__
return self._call_impl(args, kwargs)
File "C:Usersse01040434Anaconda3libsite-packagestensorflowpythoneagerfunction.py", line 1645, in _call_impl
return self._call_flat(args, self.captured_inputs, cancellation_manager)
File "C:Usersse01040434Anaconda3libsite-packagestensorflowpythoneagerfunction.py", line 1746, in _call_flat
ctx, args, cancellation_manager=cancellation_manager))
File "C:Usersse01040434Anaconda3libsite-packagestensorflowpythoneagerfunction.py", line 598, in call
ctx=ctx)
File "C:Usersse01040434Anaconda3libsite-packagestensorflowpythoneagerexecute.py", line 60, in quick_execute
inputs, attrs, num_outputs)
InvalidArgumentError:  Incompatible shapes: [672] vs. [1024]
[[node gradients/loss/dense_5_loss/vae_loss/weighted_loss/mul_grad/Mul_1 (defined at C:Usersse01040434Anaconda3libsite-packageskerasbackendtensorflow_backend.py:3009) ]] [Op:__inference_keras_scratch_graph_1515]
Function call stack:
keras_scratch_graph

Answer 1

您正在创建一个具有batch_size样本的随机张量，其中batch_size是代码中的固定预设值。然而，请注意，模型可能不一定需要batch_size输入样本那么多(例如，最后一批训练/测试数据可能具有较小数量的样本(。相反，在模型实现依赖于批大小的动态值的情况下，应该使用keras.backend.shape函数动态获取它：

def sampling(args):
# ...
eps = K.random_normal(shape=(K.shape(mu)[0], min_dim)