如何在tensorflow中累积梯度？

我有一个类似的问题

因为我的资源有限，而且我使用的是一个深度模型（VGG-16）——用于训练一个三重网络——我想为一个训练示例大小的128批累积梯度，然后传播错误并更新权重

---

我不清楚该怎么做。我使用tensorflow，但欢迎使用任何实现/伪代码

让我们浏览一下您喜欢的答案之一中提出的代码：

## Optimizer definition - nothing different from any classical example
opt = tf.train.AdamOptimizer()

## Retrieve all trainable variables you defined in your graph
tvs = tf.trainable_variables()
## Creation of a list of variables with the same shape as the trainable ones
# initialized with 0s
accum_vars = [tf.Variable(tf.zeros_like(tv.initialized_value()), trainable=False) for tv in tvs]
zero_ops = [tv.assign(tf.zeros_like(tv)) for tv in accum_vars]

## Calls the compute_gradients function of the optimizer to obtain... the list of gradients
gvs = opt.compute_gradients(rmse, tvs)

## Adds to each element from the list you initialized earlier with zeros its gradient (works because accum_vars and gvs are in the same order)
accum_ops = [accum_vars[i].assign_add(gv[0]) for i, gv in enumerate(gvs)]

## Define the training step (part with variable value update)
train_step = opt.apply_gradients([(accum_vars[i], gv[1]) for i, gv in enumerate(gvs)])

第一部分主要是在图形中添加新的

变量

和

ops

，这将允许您

使用ops

accum\u ops

in（变量列表）累积梯度

accum\u vars

使用ops

train\u步骤更新模型重量

然后，要在培训时使用它，您必须遵循以下步骤（仍然来自您链接的答案）：

---

Tensorflow 2.0兼容答案：与上文提到的Pop答案和中提供的解释一致，下面提到的是Tensorflow 2.0版中累积梯度的代码：

def train(epochs):
  for epoch in range(epochs):
    for (batch, (images, labels)) in enumerate(dataset):
       with tf.GradientTape() as tape:
        logits = mnist_model(images, training=True)
        tvs = mnist_model.trainable_variables
        accum_vars = [tf.Variable(tf.zeros_like(tv.initialized_value()), trainable=False) for tv in tvs]
        zero_ops = [tv.assign(tf.zeros_like(tv)) for tv in accum_vars]
        loss_value = loss_object(labels, logits)

       loss_history.append(loss_value.numpy().mean())
       grads = tape.gradient(loss_value, tvs)
       #print(grads[0].shape)
       #print(accum_vars[0].shape)
       accum_ops = [accum_vars[i].assign_add(grad) for i, grad in enumerate(grads)]



    optimizer.apply_gradients(zip(grads, mnist_model.trainable_variables))
    print ('Epoch {} finished'.format(epoch))

# call the above function    
train(epochs = 3)

---

完整的代码可以在这里找到。

为什么不使用您链接的问题的答案？@Pop，因为我不理解它们。我在寻找更详细的东西（初学者级别），所以您将

sess.run（train\u step）

放在循环之外。这意味着重量更新将在计算最后一批的梯度后发生，对吗？如果我们把它放在循环中，它会在每个历元之后发生，对吗？应该是

优化器。应用梯度（zip（accum\u ops，mnist\u model.trainable\u variables））

？我也无法在tf.function中创建tf.Variable，有什么建议吗？我在遵循这段代码时也遇到了问题，我发布了一个工作版本的链接问题；

def train(epochs):
  for epoch in range(epochs):
    for (batch, (images, labels)) in enumerate(dataset):
       with tf.GradientTape() as tape:
        logits = mnist_model(images, training=True)
        tvs = mnist_model.trainable_variables
        accum_vars = [tf.Variable(tf.zeros_like(tv.initialized_value()), trainable=False) for tv in tvs]
        zero_ops = [tv.assign(tf.zeros_like(tv)) for tv in accum_vars]
        loss_value = loss_object(labels, logits)

       loss_history.append(loss_value.numpy().mean())
       grads = tape.gradient(loss_value, tvs)
       #print(grads[0].shape)
       #print(accum_vars[0].shape)
       accum_ops = [accum_vars[i].assign_add(grad) for i, grad in enumerate(grads)]



    optimizer.apply_gradients(zip(grads, mnist_model.trainable_variables))
    print ('Epoch {} finished'.format(epoch))

# call the above function    
train(epochs = 3)