多个批次的 TensorFlow 平均梯度

Question

这可能与Tensorflow: How to get gradients per instance in a batch? 重复。反正我也问了，因为一直没有满意的答案，而且这里的目标有点不一样。

我有一个非常大的网络，可以安装在我的 GPU 上，但我可以提供的最大批量大小是 32。任何比这更大的都会导致 GPU 内存不足。我想使用更大的批次以获得更准确的梯度近似值。

具体而言，假设我想通过依次输入 3 批 32 来计算大量 96 的梯度。我知道的最好方法是使用Optimizer.compute_gradients() 和Optimizer.apply_gradients()。这是一个如何工作的小例子

import tensorflow as tf
import numpy as np

learn_rate = 0.1

W_init = np.array([ [1,2,3], [4,5,6], [7,8,9] ], dtype=np.float32)
x_init = np.array([ [11,12,13], [14,15,16], [17,18,19] ], dtype=np.float32)

X = tf.placeholder(dtype=np.float32, name="x")
W = tf.Variable(W_init, dtype=np.float32, name="w")
y = tf.matmul(X, W, name="y")
loss = tf.reduce_mean(y, name="loss")

opt = tf.train.GradientDescentOptimizer(learn_rate)
grad_vars_op = opt.compute_gradients(loss)

sess = tf.Session()
sess.run(tf.global_variables_initializer())

# Compute the gradients for each batch
grads_vars1 = sess.run(grad_vars_op, feed_dict = {X: x_init[None,0]})
grads_vars2 = sess.run(grad_vars_op, feed_dict = {X: x_init[None,1]})
grads_vars3 = sess.run(grad_vars_op, feed_dict = {X: x_init[None,2]})

# Separate the gradients from the variables
grads1 = [ grad for grad, var in grads_vars1 ]
grads2 = [ grad for grad, var in grads_vars2 ]
grads3 = [ grad for grad, var in grads_vars3 ]
varl   = [ var  for grad, var in grads_vars1 ]

# Average the gradients
grads  = [ (g1 + g2 + g3)/3 for g1, g2, g3 in zip(grads1, grads2, grads3)]

sess.run(opt.apply_gradients(zip(grads,varl)))

print("Weights after 1 gradient")
print(sess.run(W))

现在这一切都非常丑陋和低效，因为前向传播是在 GPU 上运行的，而平均梯度发生在 CPU 上，然后再次在 GPU 上应用它们。

此外，此代码会引发异常，因为grads 是np.arrays 的列表，要使其正常工作，必须为每个渐变创建一个tf.placeholder。

我确定应该有更好、更有效的方法来做到这一点？有什么建议吗？

Answer 1

您可以创建trainable_variables 的副本并累积批量梯度。这里有几个简单的步骤可以遵循

...
opt = tf.train.GradientDescentOptimizer(learn_rate)

# constant to scale sum of gradient
const = tf.constant(1/n_batches)
# get all trainable variables
t_vars = tf.trainable_variables()
# create a copy of all trainable variables with `0` as initial values
accum_tvars = [tf.Variable(tf.zeros_like(tv.initialized_value()),trainable=False) for t_var in t_vars]                                        
# create a op to initialize all accums vars
zero_ops = [tv.assign(tf.zeros_like(tv)) for tv in accum_tvars]

# compute gradients for a batch
batch_grads_vars = opt.compute_gradients(loss, t_vars)
# collect the (scaled by const) batch gradient into accumulated vars 
accum_ops = [accum_tvars[i].assign_add(tf.scalar_mul(const, batch_grad_var[0]) for i, batch_grad_var in enumerate(batch_grads_vars)]

# apply accums gradients 
train_step = opt.apply_gradients([(accum_tvars[i], batch_grad_var[1]) for i, batch_grad_var in enumerate(batch_grads_vars)])
# train_step = opt.apply_gradients(zip(accum_tvars, zip(*batch_grads_vars)[1])

while True:
   # initialize the accumulated gards
   sess.run(zero_ops)

   # number of batches for gradient accumulation 
   n_batches = 3
   for i in xrange(n_batches):
       sess.run(accum_ops, feed_dict={X: x_init[:, i]})

   sess.run(train_step)