v0.1.0 initial release

[EternalmBlue]

DomainPostTrain v0.1.0 is the initial public release.

Release notes: https://github.com/EternalmBlue/DomainPostTrain/releases/tag/v0.1.0

This version focuses on a domain-neutral post-training workflow:

• CPT -> Fact-SFT -> optional DPO -> merge -> quality evaluation
• Static mock data for CPT, SFT, DPO, and evaluation examples
• Qwen3.5 4B default config
• GGUF export from the merged model, with ONNX kept optional
• OpenAI-compatible inference service
• CI, issue templates, PR template, and discussion forms

Use Q&A for setup and troubleshooting, Ideas for broader proposals, and Show and tell for safe experiment reports or demos.