有没有一种方法可以手动训练Tensorflow Keras模型而不使用fit方法，或者我很害怕？

Question

我刚刚开始学习Tensorflow和Keras，我想知道是否还有另一种方法（更手动地）在不使用拟合方法的情况下训练模型。

我想手动执行优化过程：我想使用自己的损失函数自己更新参数，并用我计算出的损失更新这些参数。

有没有办法做到这一点，或者我吓坏了？

我之所以这样问，是因为我想了解更多Keras模型的工作原理，以及是否有任何方法可以在不使用fit方法的情况下训练网络，这看起来像在做很多事情。

Answer 1

请看一下本教程：Tensorflow - Custom training。解释了如何创建自己的损失函数，自定义优化器以及如何定义训练循环。您可以完全控制培训过程。

我在这里复制了您可能最感兴趣的代码：

# Custom loss
def loss(target_y, predicted_y):
  return tf.reduce_mean(tf.square(target_y - predicted_y))

# Define a training loop
def train(model, inputs, outputs, learning_rate):
  with tf.GradientTape() as t:
    current_loss = loss(outputs, model(inputs))
  dW, db = t.gradient(current_loss, [model.W, model.b])
  model.W.assign_sub(learning_rate * dW)
  model.b.assign_sub(learning_rate * db)

model = Model()

Ws, bs = [], []
epochs = range(10)
for epoch in epochs:
  Ws.append(model.W.numpy())
  bs.append(model.b.numpy())
  current_loss = loss(outputs, model(inputs))

  train(model, inputs, outputs, learning_rate=0.1)
  print('Epoch %2d: W=%1.2f b=%1.2f, loss=%2.5f' %
        (epoch, Ws[-1], bs[-1], current_loss))

该教程使用了一个非常简单的线性模型，但是对于理解这一点非常有用。但是，如果您对更复杂的东西感兴趣，请查看本教程：Tensorflow - Custom training: walkthrough

#Create the model 
model = tf.keras.Sequential([
  tf.keras.layers.Dense(10, activation=tf.nn.relu, input_shape=(4,)),  # input shape required
  tf.keras.layers.Dense(10, activation=tf.nn.relu),
  tf.keras.layers.Dense(3)
])

# You can define your own loss function here
loss_object = tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True)
def loss(model, x, y, training):
  # training=training is needed only if there are layers with different
  # behavior during training versus inference (e.g. Dropout).
  y_ = model(x, training=training)

  return loss_object(y_true=y, y_pred=y_)

# Create here your gradient and optimizor
def grad(model, inputs, targets):
  with tf.GradientTape() as tape:
    loss_value = loss(model, inputs, targets, training=True)
  return loss_value, tape.gradient(loss_value, model.trainable_variables)
optimizer = tf.keras.optimizers.SGD(learning_rate=0.01)

# Training loop 
train_loss_results = []
train_accuracy_results = []

num_epochs = 201

for epoch in range(num_epochs):
  epoch_loss_avg = tf.keras.metrics.Mean()
  epoch_accuracy = tf.keras.metrics.SparseCategoricalAccuracy()

  # Training loop - using batches of 32
  for x, y in train_dataset:
    # Optimize the model
    loss_value, grads = grad(model, x, y)
    optimizer.apply_gradients(zip(grads, model.trainable_variables))

    # Track progress
    epoch_loss_avg.update_state(loss_value)  # Add current batch loss
    # Compare predicted label to actual label
    # training=True is needed only if there are layers with different
    # behavior during training versus inference (e.g. Dropout).
    epoch_accuracy.update_state(y, model(x, training=True))

  # End epoch
  train_loss_results.append(epoch_loss_avg.result())
  train_accuracy_results.append(epoch_accuracy.result())

  if epoch % 50 == 0:
    print("Epoch {:03d}: Loss: {:.3f}, Accuracy: {:.3%}".format(epoch,
                                                                epoch_loss_avg.result(),
                                                                epoch_accuracy.result()))