tracel-ai/burn · error

int_cumprod: unsupported dtype {:?}

Error message

int_cumprod: unsupported dtype {:?}

What it means

int_cumprod computes the cumulative product along a dimension, dispatching per integer dtype (I64..U8). The panic fires when the tensor dtype is not one of these integer types (float or bool), meaning a non-integer tensor reached the int cumulative-product path. It is a dtype-dispatch guard, not a data error.

Source

Thrown at crates/burn-flex/src/ops/int.rs:710

            DType::U64 => crate::ops::cumulative::cumsum::<u64>(tensor, dim),
            DType::U32 => crate::ops::cumulative::cumsum::<u32>(tensor, dim),
            DType::U16 => crate::ops::cumulative::cumsum::<u16>(tensor, dim),
            DType::U8 => crate::ops::cumulative::cumsum::<u8>(tensor, dim),
            dt => panic!("int_cumsum: unsupported dtype {:?}", dt),
        }
    }

    fn int_cumprod(tensor: IntTensor<Flex>, dim: usize) -> IntTensor<Flex> {
        match tensor.dtype() {
            DType::I64 => crate::ops::cumulative::cumprod::<i64>(tensor, dim),
            DType::I32 => crate::ops::cumulative::cumprod::<i32>(tensor, dim),
            DType::I16 => crate::ops::cumulative::cumprod::<i16>(tensor, dim),
            DType::I8 => crate::ops::cumulative::cumprod::<i8>(tensor, dim),
            DType::U64 => crate::ops::cumulative::cumprod::<u64>(tensor, dim),
            DType::U32 => crate::ops::cumulative::cumprod::<u32>(tensor, dim),
            DType::U16 => crate::ops::cumulative::cumprod::<u16>(tensor, dim),
            DType::U8 => crate::ops::cumulative::cumprod::<u8>(tensor, dim),
            dt => panic!("int_cumprod: unsupported dtype {:?}", dt),
        }
    }

    fn int_cummin(tensor: IntTensor<Flex>, dim: usize) -> IntTensor<Flex> {
        match tensor.dtype() {
            DType::I64 => crate::ops::cumulative::cummin::<i64>(tensor, dim),
            DType::I32 => crate::ops::cumulative::cummin::<i32>(tensor, dim),
            DType::I16 => crate::ops::cumulative::cummin::<i16>(tensor, dim),
            DType::I8 => crate::ops::cumulative::cummin::<i8>(tensor, dim),
            DType::U64 => crate::ops::cumulative::cummin::<u64>(tensor, dim),
            DType::U32 => crate::ops::cumulative::cummin::<u32>(tensor, dim),
            DType::U16 => crate::ops::cumulative::cummin::<u16>(tensor, dim),
            DType::U8 => crate::ops::cumulative::cummin::<u8>(tensor, dim),
            dt => panic!("int_cummin: unsupported dtype {:?}", dt),
        }
    }

    fn int_cummax(tensor: IntTensor<Flex>, dim: usize) -> IntTensor<Flex> {

View on GitHub (pinned to d16f7ba2ed)

Solutions

1. Check the tensor dtype before the call; cast to an integer dtype (e.g. i64) if the data is logically integer.
2. Route float tensors to a float-compatible cumulative product implementation.
3. Audit the preceding operations for implicit dtype promotion and pin dtypes with explicit casts.
4. Add an upstream assert on dtype so the failure surfaces at the pipeline boundary.

Example fix

// before
let out = t.cumprod(dim); // t is F32 -> panic
// after
let out_i = t.cast::<i64>();
let out = out_i.cumprod(dim);

assert!(t.dtype().is_int(), "cumprod requires an int tensor, got {:?}", t.dtype());

fn is_int_tensor(t: &FlexTensor) -> bool { matches!(t.dtype(), DType::I64 | DType::I32 | DType::I16 | DType::I8 | DType::U64 | DType::U32 | DType::U16 | DType::U8) }

When it happens

Trigger: Calling cumprod (Tensor::cumprod / int_cumprod) on a Flex tensor with dtype F32/F64/F16/BF16 or Bool.

Common situations: Porting PyTorch code where cumprod works on any dtype; a preceding op (mul, cast, division) silently promoted the tensor to float; deserialized tensors with unexpected dtype.

Related errors

• float_cumprod: unsupported dtype {:?}
• burn-flex does not support Bool(U32) storage (only Native an
• compare_int: unsupported dtype {:?}
• compare_int_elem: unsupported dtype {:?}
• any_float: unsupported dtype {:?}

AI-assisted analysis of tracel-ai/burn@d16f7ba2ed (2026-09-05). Data as JSON: /api/errors/c6be9dbeb8a2c7b6. Report an issue: GitHub.