feat(purejava): support MiniCPM5 1B

[bsbodden]

Summary

• add the official OpenBMB MiniCPM5 1B Q4_K_M as a mandatory ModelJars fixture
• validate exact tokenizer references, explicit Q/K/V widths, 219-tensor inventory, and 131K context metadata
• match a stable four-token llama.cpp b9960 Java code-completion oracle
• add the fixture to automatic real-model CI and documentation

Test-first evidence

• registry resolution first failed with no marker
• pure-Java resolution then failed while its backend flag was false
• integration tests failed hard while the required GGUF was absent
• the verified installer downloaded the exact 688,065,920-byte artifact and validated SHA-256

Validation

• exact oracle [5028, 6706, 5018, 1735] for public static void main(String[] args) {
• full unfiltered real-model integration suite (5m30s)
• full 135-task multi-module build
• pure-Java unit, SpotBugs, and Spotless gates
• workflow YAML parse and git diff --check

Upstream

• https://huggingface.co/openbmb/MiniCPM5-1B-GGUF
• Add MiniCPM5 tokenizer support ggml-org/llama.cpp#23384

[github-advanced-security]

You are seeing this message because GitHub Code Scanning has recently been set up for this repository, or this pull request contains the workflow file for the Code Scanning tool.

What Enabling Code Scanning Means:

• The 'Security' tab will display more code scanning analysis results (e.g., for the default branch).
• Depending on your configuration and choice of analysis tool, future pull requests will be annotated with code scanning analysis results.
• You will be able to see the analysis results for the pull request's branch on this overview once the scans have completed and the checks have passed.

For more information about GitHub Code Scanning, check out the documentation.