Loss weighting
Three versions of the two-output model for alphabet and character prediction that you built before have been trained: model_a
, model_b
, and model_c
. For all three, the loss was defined as follows:
loss_alpha = criterion(outputs_alpha, labels_alpha)
loss_char = criterion(outputs_char, labels_char)
loss = ((1 - char_weight) * loss_alpha) + (char_weight * loss_char)
However, each of the three models was trained with a different char_weight
: 0.1
, 0.5
, or 0.9
.
Use the function you have defined in the previous exercise, evaluate_model()
, to check the accuracy of each model. Which char_weight
was used to train which model?
This exercise is part of the course
Intermediate Deep Learning with PyTorch
Hands-on interactive exercise
Turn theory into action with one of our interactive exercises
