keras-team/keras · error · ValueError

Invalid feature type

Error message

Invalid feature type: {feature}

What it means

FeatureSpace dispatches feature preprocessing by matching the feature-type string against a fixed set of supported names ('float', 'float_normalized', 'float_rescaled', 'float_discretized', 'integer_categorical', 'string_categorical', etc.). This error means the string in your `features` dict value matched none of them — typically a misspelling or an unsupported custom type.

Solutions

1. Use one of the supported feature type strings, e.g. 'float', 'float_normalized', 'float_rescaled', 'float_discretized', 'integer_categorical', 'string_categorical'
2. For unsupported preprocessing, pass a custom Keras layer instance instead of a string as the feature value

When it happens

Trigger: Thrown at keras/src/layers/preprocessing/feature_space.py:489 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of keras-team/keras@7a34a03db6 (2026-08-25). Data as JSON: /api/errors/17af69ce2b9ee3c9. Report an issue: GitHub.

Appendix: source

Thrown at keras/src/layers/preprocessing/feature_space.py:489

            return self.float(name=name)
        elif feature == "float_normalized":
            return self.float_normalized(name=name)
        elif feature == "float_rescaled":
            return self.float_rescaled(name=name)
        elif feature == "float_discretized":
            return self.float_discretized(
                name=name, num_bins=self.num_discretization_bins
            )
        elif feature == "integer_categorical":
            return self.integer_categorical(name=name)
        elif feature == "string_categorical":
            return self.string_categorical(name=name)
        elif feature == "integer_hashed":
            return self.integer_hashed(self.hashing_dim, name=name)
        elif feature == "string_hashed":
            return self.string_hashed(self.hashing_dim, name=name)
        else:
            raise ValueError(f"Invalid feature type: {feature}")

    def _cross_to_crosser(self, cross):
        return layers.HashedCrossing(cross.crossing_dim, name=cross.name)

    def _list_adaptable_preprocessors(self):
        adaptable_preprocessors = []
        for name in self.features.keys():
            preprocessor = self.preprocessors[name]
            # Special case: a Normalization layer with preset mean/variance.
            # Not adaptable.
            if isinstance(preprocessor, layers.Normalization):
                if preprocessor.input_mean is not None:
                    continue
            # Special case: a TextVectorization layer with provided vocabulary.
            elif isinstance(preprocessor, layers.TextVectorization):
                if preprocessor._has_input_vocabulary:
                    continue
            if hasattr(preprocessor, "adapt"):

View on GitHub (pinned to 7a34a03db6)