hankcs/HanLP · error · ValueError

Unsupported argument type: {item}

Error message

Unsupported argument type: {item}

What it means

Feedforward requires the hidden_dims list length to exactly equal num_layers, since each layer consumes one hidden dimension. If you pass, e.g., num_layers=2 with hidden_dims=[100], the config is inconsistent and the constructor raises ValueError.

Source

Thrown at hanlp/common/transform.py:87

    def __init__(self, *fields) -> None:
        super().__init__()
        for each in fields:
            self.append(FieldToIndex(each))

    def append(self, item: Union[str, Tuple[str, Vocab], Tuple[str, str, Vocab], FieldToIndex]) -> None:
        if isinstance(item, str):
            item = FieldToIndex(item)
        elif isinstance(item, (list, tuple)):
            if len(item) == 2:
                item = FieldToIndex(src=item[0], vocab=item[1])
            elif len(item) == 3:
                item = FieldToIndex(src=item[0], dst=item[1], vocab=item[2])
            else:
                raise ValueError(f'Unsupported argument length: {item}')
        elif isinstance(item, FieldToIndex):
            pass
        else:
            raise ValueError(f'Unsupported argument type: {item}')
        super(self).append(item)

    def save_vocab(self, save_dir):
        for each in self:
            each.save_vocab(save_dir, None)

    def load_vocab(self, save_dir):
        for each in self:
            each.load_vocab(save_dir, None)


class VocabDict(SerializableDict):

    def __init__(self, *args, **kwargs) -> None:
        """A dict holding :class:`hanlp.common.vocab.Vocab` instances. When used as a transform, it transforms the field
        corresponding to each :class:`hanlp.common.vocab.Vocab` into indices.

        Args:

View on GitHub (pinned to ddb1299bdd)

Solutions

1. Set num_layers = len(hidden_dims) (or extend hidden_dims to match num_layers)
2. Pass hidden_dims as an explicit list with one entry per layer, e.g. [128, 64] for 2 layers
3. Validate config programmatically before building the model

Example fix

# before
Feedforward(input_dim=300, num_layers=2, hidden_dims=[128])
# after
Feedforward(input_dim=300, num_layers=2, hidden_dims=[128, 128])

assert len(hidden_dims) == num_layers, f'hidden_dims {len(hidden_dims)} != num_layers {num_layers}'

try:
    ff = Feedforward(input_dim, num_layers, hidden_dims)
except ValueError as e:
    raise ValueError(f'Invalid Feedforward config: {e}') from e

When it happens

Trigger: Constructing Feedforward(input_dim, num_layers, hidden_dims, ...) where len(hidden_dims) != num_layers, or passing a single int where a per-layer list is expected.

Common situations: Typos in component configs (e.g. transformer/tagger head configs); refactoring a 1-layer FFN to N layers and forgetting to extend hidden_dims; passing hidden_dims as a scalar instead of a list.

Understand the failure class

Background: Config validation failed: what "invalid value for {key}" and settings-rejection errors mean across 19 open-source libraries — this error's family across 19 libraries.

Related errors

• Unrecognized mapper type {mapper}
• self.model.config.pad_token_id has to be defined.
• embed_dim must be divisible by num_heads (got `embed_dim`: {
• Attention weights should be of size {(bsz * self.num_heads,
• You cannot specify both input_ids and inputs_embeds at the s

AI-assisted analysis of hankcs/HanLP@ddb1299bdd (2026-08-27). Data as JSON: /api/errors/5e1a47c6a430a7ce. Report an issue: GitHub.