Misleading `Tensor::matmul` documentation

First off, great job on this project!

I noticed that Gemm (at least for cuda from what I've seen) is not supported for tensors with dims > 4, however, the docs here seem to describe a tensor with arbitrary dimensions https://docs.rs/candle-core/latest/candle_core/struct.Tensor.html#method.matmul. It would be helpful to document this limitation.

Start with the Tensor::matmul API documentation at the linked docs.rs entry and locate its corresponding Rust documentation source. Document that the CUDA Gemm implementation does not support tensors with dimensions greater than four, then verify that the generated API documentation states the limitation accurately.

Written by the indexing model from the issue text.