[docs] Deploying

[stevhliu]

This PR focuses on several improvements to the serving and deploying docs:

• consolidates integration guides and ExecuTorch/ONNX export guides to reduce navigation overhead and group related content together
• improve scope of the serving.md doc to only the transformers serve CLI. clearer intro on when to use it vs optimized inference engines, more complete endpoint descriptions (v1/audio/transcriptions and v1/models), MCP example with GenerationConfig, port forwarding, and better structure for optimizations

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.