marigold_ensemble integer fix

[graemeniedermayer]

For the integer issue #388

This is probably the best place to cast to integers. Feel free to modify any other details.

---

The float16 does seem to be working. The vram significantly lower, although I keep getting an erroneous error message.

Pipelines loaded with `torch_dtype=torch.float16` cannot run with `cpu` device. It is not recommended to move them to `cpu` as running them will fail. Please make sure to use an accelerator to run the pipeline in inference, due to the lack of support for`float16` operations on this device in PyTorch. Please, remove the `torch_dtype=torch.float16` argument, or use another device for inference.

Hopefully this will be fixed by future marigold updates.

[semjon00]

More of less ok. You should check existence in the list first. What if at some point it turns out that not all values are always passed?