docs: fix outdated usage of llama-simple

[rand-fly]

• I have read the contributing guidelines
• Self-reported review complexity:
  • Low
  • Medium
  • High

Since #9803, the command-line interface for llama-simple has been updated from the common format to -m model.gguf [-n n_predict] [-ngl n_gpu_layers] [prompt]. However, the documentation has not been updated to reflect this change.

The documentation "Building on Android" is also impacted, but it might be better to just replace llama-simple with the more sophisticated llama-cli.