---

name: lora_reduced_training description: Run a bounded LoRA-only optimization loop and log mechanism-faithful training evidence.

Lora Reduced Training

When to use

Use this skill when full model training is blocked but a reduced, mechanism-faithful LoRA optimizer step is needed. It consumes tiny supervised examples, updates only A/B, records before-after parameters, and verifies frozen base weights.

Inputs

• Base matrix or named parameters relevant to the module.
• LoRA rank r, scaling alpha, and small task examples when training or evaluation is needed.

Outputs

• Deterministic JSON-serializable mechanism checks.
• Updated LoRA parameters or validation metrics, depending on the module.

Workflow

1. Keep pretrained/base weights immutable unless explicitly performing a merge copy for inference.
2. Represent the adaptation as delta_W = (alpha / r) * B @ A.
3. Verify that only LoRA matrices are treated as task-specific trainable parameters.
4. For recovery, log rank, loss before and after, parameter changes, and merge equivalence.

Validation

Run python -m pytest tests or validate the skill tree with the Distiller module-to-skill validator. The included tests use only the Python standard library.

Limitations

This skill provides the reusable mechanism; full paper benchmark scores still require real pretrained models, datasets, and training infrastructure.