Chapter 19 - Custom Training Loop Model Compile

When running custom training loop I get:

RuntimeError: You must compile your model before training/testing. Use `model.compile(optimizer, loss)`.

Inserting:

model.compile(loss="sparse_categorical_crossentropy",
optimizer=keras.optimizers.SGD(lr=1e-2),
metrics=["accuracy"])

before calling model.fit resolves but i'm not sure if this is correct.