public compile(optimizer as String = "adam", loss as String = "sparse_categorical_crossentropy", metrics as IList<String> = null) as Void
Resolves standard optimizer/loss names to actual typed implementations. sparse_categorical_crossentropy consumes probabilities in this overload; use an explicit loss with fromLogits=true for a linear final layer.
| Parameter | Description |
|---|---|
optimizer | adam (rate 0.001) or sgd (rate 0.01). |
loss | sparse_categorical_crossentropy, mean_squared_error or mse. |
metrics | Null selects accuracy for classification and no metric for regression. Explicit regression metrics must be empty. |
model.compile(optimizer="adam", loss="sparse_categorical_crossentropy", metrics=["accuracy"])