训练模型时，Keras输出数据中显示的90/90是什么? [英] what is the 90/90 shown in Keras output data while training the model?

问题描述

90/90 - 0s - loss: 0.0098 - accuracy: 0.0011这是训练模型时的输出. 那90/90是多少?如果是通过的样本数，则我的数据集有1001行和9列！所以输出不应该是1001/1001吗? 我的代码如下所示:

90/90 - 0s - loss: 0.0098 - accuracy: 0.0011 This is my output while training my model . what is that 90/90 ? if it is the number of samples passed , My data set has 1001 rows and 9 columns ! so shouldn't the output be 1001/1001 ? My code is shown below :

model = Sequential([
Dense(9,input_shape=(9,),activation='relu'),
Dense(25,activation='relu'),
Dense(18,activation='relu'),
Dense(1,activation='sigmoid')
])
model.compile(
optimizer=Adam(learning_rate=0.001),
loss='mean_squared_error',
metrics=['accuracy']
)
model.fit(
scaled_train_data,
label_set,
validation_split=0.1,
batch_size=10,
epochs=2000,
shuffle=True,
verbose=2
)

推荐答案

您的验证拆分为0.1，因此训练有900个样本.因此，使用batch_size=10时，有90个批次.这就是90/90的原因.

You have a validation split of 0.1, so training has 900 samples. Thus, with batch_size=10, 90 batches are there. That is the reason of 90/90.

这篇关于训练模型时，Keras输出数据中显示的90/90是什么?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！