Mask for the table. Indicates which tokens belong to the table (1). Question tokens, table headers and
            padding are 0.
        labels (`tf.Tensor` of shape `(batch_size, seq_length)`, *optional*):
            Labels per token for computing the hierarchical cell selection loss. This encodes the positions of the
            answer appearing in the table. Can be obtained using [`AutoTokenizer`].

            - 1 for tokens that are **part of the answer**,
            - 0 for tokens that are **not part of the answer**.

        aggregation_labels (`tf.Tensor` of shape `(batch_size, )`, *optional*):
            Aggregation function index for every example in the batch for computing the aggregation loss. Indices
            should be in `[0, ..., config.num_aggregation_labels - 1]`. Only required in case of strong supervision for
            aggregation (WikiSQL-supervised).
        float_answer (`tf.Tensor` of shape `(batch_size, )`, *optional*):
            Float answer for every example in the batch. Set to *float('nan')* for cell selection questions. Only
            required in case of weak supervision (WTQ) to calculate the aggregate mask and regression loss.
        numeric_values (`tf.Tensor` of shape `(batch_size, seq_length)`, *optional*):
            Numeric values of every token, NaN for tokens which are not numeric values. Can be obtained using
            [`AutoTokenizer`]. Only required in case of weak supervision for aggregation (WTQ) to calculate the
            regression loss.
        numeric_values_scale (`tf.Tensor` of shape `(batch_size, seq_length)`, *optional*):
            Scale of the numeric values of every token. Can be obtained using [`AutoTokenizer`]. Only required in case
            of weak supervision for aggregation (WTQ) to calculate the regression loss.

        Returns:

        Examples:

        ```python
        >>> from transformers import AutoTokenizer, TapasForQuestionAnswering
        >>> import pandas as pd

        >>> tokenizer = AutoTokenizer.from_pretrained("google/tapas-base-finetuned-wtq")
        >>> model = TapasForQuestionAnswering.from_pretrained("google/tapas-base-finetuned-wtq")

        >>> data = {
        ...     "Actors": ["Brad Pitt", "Leonardo Di Caprio", "George Clooney"],
        ...     "Age": ["56", "45", "59"],
        ...     "Number of movies": ["87", "53", "69"],
        ... }
        >>> table = pd.DataFrame.from_dict(data)
        >>> queries = ["How many movies has George Clooney played in?", "How old is Brad Pitt?"]

        >>> inputs = tokenizer(table=table, queries=queries, padding="max_length", return_tensors="tf")
        >>> outputs = model(**inputs)

        >>> logits = outputs.logits
        >>> logits_aggregation = outputs.logits_aggregation
        ```r5