Can I change the dimension of the embeddings to 128 instead of 192? Should I changelin_neurons:192 to 128 in hyperparams.yaml and train it again? Will the accuracy be the same?
· Sign up or log in to comment