docs: Add Generative LLM support for INFERENCE data event

[javoweb-fc]

Description

This PR extends the INFERENCE data event documentation (stacked on #79) to add support for on-device Generative LLM inference, alongside the existing Vision ML modes.

Key Changes

1. Execution Modes: Adds Generative LLM as a supported execution mode, alongside Vision ML and Legacy Vision ML (ONNX).
2. Model Detection: Adds .gguf, .litertlm, and .task file extensions as Generative LLM model types.
3. Parameters: Documents the new config schema for Generative LLM (prompt, systemPrompt, temperature, topK, topP, maxTokens, contextSize, stopTokens).
4. Callback Signature: Documents the LLM output shape (result.outputs.text, result.modelType).
5. Device Warnings: Adds LLM-specific battery/memory guidance.
6. New Example: Adds an on-device LLM text summarization example (gemma-4-e2b.litertlm).

JIRA / Issue Reference

Ref: FLCRM-20635

Notes

Part 2 of 2 splitting the original #76. Stacked on top of #79 — base branch should be updated to v2 once #79 merges.

[Copilot]

Pull request overview

Note

Copilot couldn't run its full agentic review because no GitHub Actions runner was available. Make sure your repository has a runner available to run Copilot's review, or add a copilot-setup-steps.yml file specifying one with the runs-on attribute. See the docs for more details.

Extends the INFERENCE data event documentation to cover on-device Generative LLM inference (in addition to existing Vision ML) including configuration parameters, output shape, and a new example.

Changes:

• Adds “Generative LLM” as a documented execution mode with an LLM-specific config schema.
• Updates model auto-detection docs and examples to include LLM model file extensions and outputs.
• Adds an on-device LLM summarization example and additional device resource guidance.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.