Target
Make a correct training recipe converge stably within Mac memory and time bounds.
Failure this recipe addresses
Loss spikes, under-updating, or optimizer memory blocks a frozen recipe.
Data contract
The same tiny-overfit and held-out splits as the incumbent.
Method or policy
Change exactly one optimizer, learning-rate schedule, or stability term per comparison.
Evaluation contract
Convergence, target quality, wall time, and peak memory versus incumbent.
- early-step stability
- final task score
- memory
- step time
Budget and stop rule
Fresh experiment only; one bounded A/B before a sweep.
Stop if the change does not address the diagnosed bottleneck.
Decision rule
Prefer the simplest optimizer meeting stability, quality, memory, and speed gates.
Learning exercise
Diagnose three recorded loss curves as learning-rate, capacity, or data failures.
Choose an optimizer change from evidence rather than novelty.
How to read this dossier
This canonical page separates the retained repository decision from the material that informed it. A source may describe an external claim, historical measurement, local observation, or planned procedure; those evidence classes are not interchangeable. Follow the provenance links before reusing a number or method.
Status and disposition describe what PostTrainLLM retained when this record was indexed. They are not a live product promise, a newly run benchmark, or permission to restart historical work. Unknown or unmeasured fields remain unknown rather than being treated as zero.
To reuse the record, first name the exact claim you need and trace it to the linked source. Then check whether the original environment, model revision, data split, evaluator, hardware, and budget match the proposed use. If they do not, treat the record as a hypothesis or design reference and run the smallest fresh comparison that can falsify it. Preserve negative outcomes and regressions beside any improvement; a local win on one slice does not silently become a general capability claim.
Source provenance
The normalized record comes from docs/recipes/registry.json. The links below are the tracked evidence and explanatory sources preserved with the record.