tracel-ai/burn · error

Should be float, got bool

Error message

Should be float, got bool

What it means

BackendTensor::float() panics with this message when called on a Bool tensor. The method only unwraps the Float variant; boolean tensors (from comparisons/logical ops) must be converted to a numeric dtype before float extraction, so receiving one indicates a dtype-handling bug.

Source

Thrown at crates/burn-dispatch/src/tensor.rs:42

    Float(B::FloatTensorPrimitive),
    /// Int tensor handle.
    Int(B::IntTensorPrimitive),
    /// Bool tensor handle.
    Bool(B::BoolTensorPrimitive),
    /// Quantized tensor handle.
    Quantized(B::QuantizedTensorPrimitive),
    #[cfg(feature = "autodiff")]
    /// Autodiff float tensor handle.
    Autodiff(FloatTensor<Autodiff<B>>),
}

impl<B: Backend> BackendTensor<B> {
    /// Returns the inner float tensor primitive.
    pub fn float(self) -> B::FloatTensorPrimitive {
        match self {
            BackendTensor::Float(tensor) => tensor,
            BackendTensor::Int(_) => panic!("Should be float, got int"),
            BackendTensor::Bool(_) => panic!("Should be float, got bool"),
            BackendTensor::Quantized(_) => panic!("Should be float, got quantized"),
            #[cfg(feature = "autodiff")]
            BackendTensor::Autodiff(_) => panic!("Should be float, got autodiff"),
        }
    }
    /// Returns the inner float tensor primitive.
    pub fn as_float(&self) -> &B::FloatTensorPrimitive {
        match self {
            BackendTensor::Float(tensor) => tensor,
            BackendTensor::Int(_) => panic!("Should be float, got int"),
            BackendTensor::Bool(_) => panic!("Should be float, got bool"),
            BackendTensor::Quantized(_) => panic!("Should be float, got quantized"),
            #[cfg(feature = "autodiff")]
            BackendTensor::Autodiff(_) => panic!("Should be float, got autodiff"),
        }
    }

    /// Returns the inner int tensor primitive.

View on GitHub (pinned to d16f7ba2ed)

Solutions

1. Cast the Bool tensor to float first (bool -> {0.0, 1.0} cast op) before calling .float().
2. Match on the BackendTensor variant and convert Bool explicitly with a numeric cast.
3. If the tensor should never be bool, fix the upstream op producing it (e.g., use a numeric comparison result instead of a mask).
4. Use checked accessors like as_float() with a variant check for mixed-dtype paths.

Example fix

// before
let f = mask_tensor.float(); // panics: Bool
// after
let f = BackendTensor::Float(bool_to_float_cast::<B>(mask_bool)).float(); // or mask_bool.cast::<f32>() upstream

// before calling .float()
if let BackendTensor::Bool(_) = &tensor {
    // cast bool to float first
}

fn is_bool_tensor<B: Backend>(t: &BackendTensor<B>) -> bool {
    matches!(t, BackendTensor::Bool(_))
}

// guard rather than catch the panic
let f = match tensor {
    BackendTensor::Bool(b) => BackendTensor::Float(bool_cast::<B>(b)),
    BackendTensor::Float(f) => f,
    other => bail!("expected float or bool, got {other:?}"),
};

When it happens

Trigger: Calling BackendTensor::float() on a tensor holding BackendTensor::Bool(_), typically the result of comparison (==, <, >), logical ops, or mask creation flowing into float-only code.

Common situations: Using boolean masks as if they were float tensors (e.g., multiplying without casting); APIs whose signatures changed to accept generic BackendTensor where callers pass masks directly.

Related errors

• Should be float, got int
• Should be float, got quantized
• Should be float, got autodiff
• float_into_int: unsupported source dtype {:?}
• float_gather: unsupported dtype {:?}

AI-assisted analysis of tracel-ai/burn@d16f7ba2ed (2026-09-05). Data as JSON: /api/errors/50908bf62df5f975. Report an issue: GitHub.