在 Keras 中绘制模型的损失和准确性

Question

我有一个函数可以在 keras 中构建模型，如下所示：

def build_model(lr = 0.0):
inp = Input(shape = (max_len,))
x = Embedding_layer
y = LSTM_layer(x)
y = Convolution_layer(y)
x = GlobalMaxPooling1D(y)
x = Dense(3, activation = "sigmoid")(x)
model = Model(inputs = inp, outputs = x)
model.compile(loss = "binary_crossentropy", optimizer = Adam(), metrics = ["accuracy"])
history = model.fit(X_train, Y_train, batch_size = 256, epochs = 3, 
verbose = 1, callbacks = [ra_val, check_point, early_stop])
model = load_model(file_path)
return model
model = build_model(lr = 1e-3)

现在我想在训练阶段后绘制历史损失和准确性，但模型没有历史选项。

如何绘制损失和精度？

Answer 1

fit函数将返回一个包含损失值和训练指标的History对象。