SLM packer command surface

Enumerates fixture generation, structural validation, manifest operations, source conversion, evaluation templates, and quality-gate commands as one operational API.

Experimental
Last verified
2026-06-25 00:00 UTC
Updated
Reading time
1 minutes

Enumerates fixture generation, structural validation, manifest operations, source conversion, evaluation templates, and quality-gate commands as one operational API.

Implementation evidence: this topic is grounded in the reviewed GGUF.MiRust.com source snapshot. It documents observed code and artifacts without claiming broad deployment, model quality, or production readiness.

Artifact generation

Direct output modes create tiny f32, tied, q8, q4, and BPE fixtures plus TinyLM-16M-shaped f32, q8, and q4 artifacts.

Admission commands

  • validate <input.slm>
  • manifest <input.slm> <output.manifest> [source-kind]
  • validate-manifest <input.slm> <input.manifest>
  • quality-gate ... runtime-smoke|assistant-quality

Source conversion commands

  • source-template
  • validate-source
  • convert-trained ... f32|q8_0|q4_0

Operational rule

The CLI and all sidecar schemas are versioned interfaces. Release automation should invoke them with absolute source identities and retain stdout, stderr, exit code, and output checksums.

Scope

This starter page defines the questions, boundaries, evidence, and failure modes that should be recorded before a capability is presented as supported.

Engineering considerations

  • Identify the source, version, target environment, and owner.
  • Separate observed values from estimates and externally reported values.
  • Record trade-offs, unsupported cases, and fallback behavior.
  • Link performance statements to a compatible benchmark methodology.

Verification questions

  • What exact artifact, revision, backend, and environment were reviewed?
  • Which assumptions could change the result?
  • Which data should be retained so another engineer can reproduce the conclusion?