[Feature] VLM eagle3 training and release

[jiapingW]

Checklist

• 1. If the issue you raised is not a feature but a question, please raise a discussion at https://github.com/sgl-project/SpecForge/discussions/new/choose Otherwise, it will be closed.
• 2. Please use English, otherwise it will be closed.

Motivation

To improve the inference speed of VLM, it is necessary to train an Eagle3 model for VLM. Specforge requires an Eagle3 training model capable of running various VLM models such as Qwen2.5-VL, Qwen3-VL, and GLM-4.6V series and so on. We also need to release some community-available VLM Eagle3 models.

Related resources

No response