Improve documentation and perhaps warn users about GPU requirements

[ns-rse]

I find I get the following warnings about CUDA as I don't have the correct nVidia driver installed.

❱ topostats --help        
2024-10-11 15:14:19.949079: I external/local_xla/xla/tsl/cuda/cudart_stub.cc:32] Could not find cuda drivers on your machine, GPU will not be used.
2024-10-11 15:14:20.063398: I external/local_xla/xla/tsl/cuda/cudart_stub.cc:32] Could not find cuda drivers on your machine, GPU will not be used.
2024-10-11 15:14:20.253694: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:485] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
2024-10-11 15:14:20.398203: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:8454] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
2024-10-11 15:14:20.439801: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1452] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
2024-10-11 15:14:20.718118: I tensorflow/core/platform/cpu_feature_guard.cc:210] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
To enable the following instructions: AVX2 FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
2024-10-11 15:14:22.068608: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT

Users may be concerned/worried/confused by this and we should add to our documentation an explanation of why GPUs are now required and perhaps even capture these messages and replace with a custom one if possible.

[ns-rse]

We've suppressed the warnings (see #976), but and I suggested in #916 that we may want to make it optional to install TensorFlow as it may well be overkill for some users. This should of course be documented.