Add HuggingFace integration doc page

[andrewor14]

We have existing HF integrations today but they're hidden in various places:

• https://docs.pytorch.org/ao/main/serving.html
• https://huggingface.co/pytorch/Phi-4-mini-instruct-INT8-INT4

Instead, we should have dedicated pages about this integration just like our vLLM page and in our partner repos:

• https://docs.pytorch.org/ao/main/torchao_vllm_integration.html
• https://huggingface.co/docs/transformers/en/quantization/torchao
• https://huggingface.co/docs/diffusers/en/quantization/torchao
• https://docs.vllm.ai/en/latest/features/quantization/torchao.html

It should highlight:

• Example integration code snippets with HF transformers and diffusers
• Some results so far, e.g. https://github.com/sayakpaul/diffusers-torchao/blob/main/README.md (Note: diffusers-torchao is outdated, the real integration lives in diffusers now)
• The latest quantization configs we support (e.g. Float8DynamicActivationInt4WeightConfig)
• Link to serving.html for full end-to-end tutorial including inference with vLLM
• Link to torchao pages on transformers and diffusers docs

Thanks for the suggestion @sayakpaul

[sayakpaul]

Thanks, Andrew, for tracking this!

Please mention me and @stevhliu for any PR related to this effort. Happy to help review :)