Eagle3 demo

[mzegla]

No description provided.

[Copilot]

Pull request overview

This PR adds support for EAGLE3 speculative decoding in the OpenVINO Model Server demos by introducing dedicated documentation, configuration updates, and model export enhancements.

Changes:

• Added comprehensive EAGLE3 setup documentation with model preparation and deployment instructions
• Added datasets dependency to support EAGLE3 model export requirements
• Enhanced export script with EAGLE3-specific flags and draft device configuration

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 3 comments.

File	Description
demos/continuous_batching/speculative_decoding/README.md	Added EAGLE3 section with model preparation steps and deployment instructions for Qwen3-8B models
demos/common/export_models/requirements.txt	Added datasets package dependency for EAGLE3 support
demos/common/export_models/export_model.py	Added --draft_eagle3 flag and draft device configuration in template

---

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

[Copilot]

The format string placeholders are misaligned with the arguments. The order should be: draft_source_model, precision, draft_llm_model_path, and additional_options should be inserted before draft_llm_model_path in the command string, not as a separate placeholder.

Suggested change

	optimum_command = "optimum-cli export openvino --model {} --weight-format {} --trust-remote-code {} {}".format(draft_source_model, precision, additional_options, draft_llm_model_path)
	optimum_command = (
	"optimum-cli export openvino "
	f"--model {draft_source_model} "
	f"--weight-format {precision} "
	"--trust-remote-code"
	f"{additional_options} "
	f"{draft_llm_model_path}"
	)