Metal error with Flux image diffusion cast_bf16_f32 does not exist

Running the new Flux model on Mac M1 Metal

running the example
```
cargo run --features metal --example flux -r -- \
--height 1024 --width 1024 \
--prompt "a rusty robot walking on a beach holding a small torch, the robot has the word "rust" written on it, high quality, 4k" --model schnell
```

and after downloading the models and trying to run
```
Tensor[[1, 768], f32, metal:4294969333]
Error: Metal error Error while loading function: "Function 'cast_bf16_f32' does not exist"

Caused by:
Error while loading function: "Function 'cast_bf16_f32' does not exist"

```

Start with the `flux` example and the provided `cargo run --features metal --example flux` command on a Mac M1. Trace the Metal execution path that requests `cast_bf16_f32`; the work is done when the example runs without the missing-function error.

Written by the indexing model from the issue text.