Question: During the training phase, if we assume Stochastic Gradient Descent (SGD) is used and batches are processed in sequence when would the validation loss be
During the training phase, if we assume Stochastic Gradient Descent (SGD) is used and batches are processed in sequence when would the validation loss be computed, considering we have 1000 samples with 100 samples per batch
Step by Step Solution
There are 3 Steps involved in it
Get step-by-step solutions from verified subject matter experts
