Beartype and jaxtyping incorrectly catching Tensor type errors at runtime
I’m using torch, jaxtyping and beartype to type annotate my functions (with runtime type checking). I have used the @jaxtyped(typechecker=beartype) decorator.
I’m using torch, jaxtyping and beartype to type annotate my functions (with runtime type checking). I have used the @jaxtyped(typechecker=beartype) decorator.