Python：随机梯度下降的代码是如何工作的？

Question

n_epochs = 50
t0, t1 = 5, 50 # learning schedule hyperparameters
def learning_schedule(t):
return t0 / (t + t1)
theta = np.random.randn(2,1) # random initialization

for epoch in range(n_epochs):
   for i in range(m):
      random_index = np.random.randint(m)
      xi = X_b[random_index:random_index+1]
      yi = y[random_index:random_index+1]
      gradients = 2 * xi.T.dot(xi.dot(theta) - yi)
      eta = learning_schedule(epoch * m + i)
      theta = theta - eta * gradients

By convention we iterate by rounds of m iterations; each round is called an epoch. While the Batch Gradient Descent code iterated 1,000 times through the whole training set, this code goes through the training set only 50 times and reaches a pretty good solution

The author says that the code iterates through the training set only 50 times. How is that possible?

Here isn't it the case that for every epoch, 1->50, i is going from 1->100 and the training data is being iterated 50*100=5000 times?

Answer 1

It will iterate all data points 50 times for ex- If data point =100 and epoch = 50 then answer is 50*100=5000

If its Stochastic gradient descent then total iteration is 5000

If its simple Gradient descent then total itertion is 50

if you want to know about the Different gradient descent techniques visit HERE