xai-org/x-algorithm · error · ValueError

valid_block_upper and valid_block_lower must both be provide

Error message

valid_block_upper and valid_block_lower must both be provided or both be None

What it means

valid_block_upper and valid_block_lower are a paired option: they must be supplied together or both omitted. Supplying only one is ambiguous because the kernel uses them as a symmetric bound pair for partial-block validity.

Source

Thrown at phoenix/xrex/cutedsl/ranker_fa4/block_sparsity.py:479

    metadata_block_shape = expected_count_shape
    valid_block_upper = _check_and_expand_metadata_tensor(
        "valid_block_upper",
        tensors.valid_block_upper,
        metadata_block_shape,
        context,
        hint,
        mask_cnt.device,
    )
    valid_block_lower = _check_and_expand_metadata_tensor(
        "valid_block_lower",
        tensors.valid_block_lower,
        metadata_block_shape,
        context,
        hint,
        mask_cnt.device,
    )
    if (valid_block_upper is None) != (valid_block_lower is None):
        raise ValueError(
            "valid_block_upper and valid_block_lower must both be provided or both be None"
        )
    spt = tensors.spt
    if spt is not None and not isinstance(spt, bool):
        raise ValueError("spt must be a bool when provided")
    if spt is not None and dq_write_order is None:
        raise ValueError("spt requires dq_write_order to be provided")

    return BlockSparseTensorsTorch(
        mask_block_cnt=mask_cnt,
        mask_block_idx=mask_idx,
        full_block_cnt=full_cnt,
        full_block_idx=full_idx,
        cu_total_m_blocks=tensors.cu_total_m_blocks,
        cu_block_idx_offsets=tensors.cu_block_idx_offsets,
        block_size=tensors.block_size,
        dq_write_order=dq_write_order,
        dq_write_order_full=dq_write_order_full,

View on GitHub (pinned to 24c60942c5)

Solutions

1. Set both fields or neither
2. If your config dict may lack one key, default both to None together: cfg.get('valid_block_upper') and cfg.get('valid_block_lower')
3. Add a paired-presence check in your config loader

Example fix

# before
tensors = BlockSparseTensorsTorch(..., valid_block_upper=ub, valid_block_lower=None)

# after
tensors = BlockSparseTensorsTorch(..., valid_block_upper=ub, valid_block_lower=lb)
# or omit both

assert (tensors.valid_block_upper is None) == (tensors.valid_block_lower is None)

def valid_bounds_paired(t) -> bool:
    return (t.valid_block_upper is None) == (t.valid_block_lower is None)

When it happens

Trigger: Constructing BlockSparseTensorsTorch with valid_block_upper set but valid_block_lower=None (or vice versa), typically when conditionally filling one field in builder code.

Common situations: Copy-paste from an example that set both, with one line later removed; defaults from a dataclass where one field was overridden; partial deserialization from a checkpoint/config dict missing one key.

Understand the failure class

Background: Missing required parameter errors: what 'X is required' and 'the required X param is missing' mean, and how to fix them — this error's family across 27 libraries.

Related errors

• {name}_block_cnt and {name}_block_idx must be on the same de
• {name} must have dtype torch.int32
• mask_block_cnt and mask_block_idx must be provided for block
• type checking expression %s failed: invalid argument type: %
• Non-optional parameter %s must be declared before optional p

AI-assisted analysis of xai-org/x-algorithm@24c60942c5 (2026-08-28). Data as JSON: /api/errors/b78d17b6e4737143. Report an issue: GitHub.