Describe the bug
This is not actually clear whether this is a bug or a feature/documentation request but here it goes:

Running rust nightly 2024-05-30, no matter how I set up libtorch I will end up with

2024-06-08T16:45:38.148370Z ERROR burn_train::learner::application_logger: PANIC => panicked at /home/user/.cargo/registry/src/index.crates.io-6f17d22bba15001f/tch-0.15.0/src/wrappers/tensor_generated.rs:7988:40:                  
called `Result::unwrap()` on an `Err` value: Torch("Could not run 'aten::empty.memory_format' with arguments from the 'CUDA' backend. This could be because the operator doesn't exist for this backend, or was omitted during the selective....

The reason I am reporting is that this is at least the third time that I encounter this same issue for different reasons such as :

• Running rust stable (a while ago nightly was required by burn)
• Running rust nightly
• Running wrong cuda version
• Running wrong libtorch version
• Wrong environment variables setup

What is my point
I think this error is totally unhelpful and there is a loot of room for improvement regarding the setup tch-gpu.

What are you thinking ?

Should we :

1. Implement pre-flight checks
2. Improve and consolidate documentation
3. Improve the error message, reading "operator does not exist" does not hint that well at where the issue is IMHO.