[Bot] Update Inference Providers documentation

[HuggingFaceInfra]

This PR automatically upgrades the @huggingface/tasks and @huggingface/inference packages and regenerates the Inference Providers documentation by running:

cd scripts/inference-providers
pnpm update @huggingface/tasks@latest @huggingface/inference@latest
pnpm run generate

This PR was automatically created (and merged) by the Update Inference Providers Documentation workflow.

---

Note

Low Risk
Documentation-only, script-generated markdown with no runtime or API behavior changes in this repo.

Overview
Regenerated Inference Providers docs so InferenceSnippet providersMapping values match the latest warm-model catalog from @huggingface/tasks / @huggingface/inference.

VLM / chat-completion examples now point several providers at moonshotai/Kimi-K2.7-Code (replacing models such as stepfun-ai/Step-3.7-Flash, moonshotai/Kimi-K2.6, and Together’s google/gemma-4-31B-it with MiniMaxAI/MiniMax-M3). Fireworks’ provider model id updates to kimi-k2p7-code.

HF Inference task snippets swap example models for ASR (openai/whisper-large-v3-turbo), feature extraction (microsoft/harrier-oss-v1-0.6b), and translation (google-t5/t5-small). Matching task pages (automatic-speech-recognition, feature-extraction, image-text-to-text, chat-completion, translation) and provider pages (featherless-ai, fireworks-ai, hf-inference, together) are updated in lockstep.

^{Reviewed by Cursor Bugbot for commit 67a04eb. Bugbot is set up for automated code reviews on this repo. Configure here.}

[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.