v0.1.1

What's Changed

• fix: guard against list-valued attention head counts in config.json extractor by @duan-li in #16

Full Changelog: v0.1.0...v0.1.1

v0.1.0

What's Changed

• Feature/mvp by @duan-li in #1
• M2 add more extractors by @duan-li in #2
• M3 add support on quantization and merge by @duan-li in #3
• m4 batch extraction by @duan-li in #4
• add doc by @duan-li in #6
• Feature/schema review by @duan-li in #7
• feat: M5 help function — CLI explain/completion + consumer helper API by @duan-li in #8
• fix: detect sharded GGUF files instead of always labeling single by @duan-li in #9
• feat: aggregate tensors across sharded GGUF (gguf-split) parts by @duan-li in #10
• feat: split extract into --download-only / --analysis-only by @duan-li in #11
• docs: cover --download-only/--analysis-only in README + dev/pipeline docs by @duan-li in #12
• feat: add modelspec verify for schema validation of ModelSpec files by @duan-li in #13
• chore: derive version from git tags, add release workflow by @duan-li in #14
• feat: annotate the exported JSON Schema with description + version by @duan-li in #15

New Contributors

• @duan-li made their first contribution in #1

Full Changelog: https://github.com/parapulseio/ModelSpec/commits/v0.1.0